Epstein Data Wiki

The Emanuel Ungaro information memorandum, January 2011

Properties2
tagstopic, fashion, corporate-transactions, paris, due-diligence, epstein-correspondence, 2011
aliasesEmanuel Ungaro, Emanuel Ungaro SAS, Ungaro, Dossin Advisory

66 min read

The objectA 34-sheet slide-format book headed “INFORMATION MEMORANDUM”, for the Paris couture house Emanuel Ungaro, carrying the wordmark “emanuel ungaro” above the letterspaced word “P A R I S” and the footer “DOSSIN ADVISORY” on almost every sheet. It stands in this release under two identifiers, EFTA00521855 (8 sheets) and EFTA00521863 (26 sheets), whose Bates numbers run continuously from EFTA00521855 to EFTA00521888123
What it offeredNot the house. In its own words it was “being delivered to a limited number of parties, who have already expressed a preliminary interest, in order to ascertain their interest in acquiring part of the share capital of the Company”, and “It does not constitute an offer or invitation to purchase any shares, and should in no way be perceived as investment advice, recommendations or services, and can not be construed as such offer, invitation, advice, recommendation or service in any jurisdiction where it would be illegal…”4
Its own field stateIt calls itself a “draft Memorandum”4. Its cover bears no date and no addressee1. No sheet of it carries a signature, an execution block or a price, and there is no instrument in this episode for anyone to sign5
Who prepared itPrepared by the Company and circulated through the advisory firm named in the footer of almost every sheet, Dossin Advisory. The memorandum’s legal section names that firm’s principal by name thirteen times, and his name stands in the legal-notice passages this page quotes; this page’s own prose refers to him by office. See Who is named here, and who is not467
Who owned the companyAsim Abdullah. The memorandum’s own ownership slide reads “‘65-‘95 : Brand owned and managed by Mr Ungaro”, “‘95-‘05 : Ownership transferred to the Ferragamo family”, “Since ‘05 : Ownership transferred to Mr Asim Abdullah”8
How it reached EpsteinBy an introduction from David J Mitchell on 21 January 2011, and by electronic forwarding on 23 January 2011 through an intermediary this page does not name91011
What Epstein didHe asked a third party about the house’s new designer; he recorded that the owner was coming to see him; and he wrote back rejecting the book and listing what it would have to contain instead121314
Was he retained, paid or mandated?Nothing in the material read for this page shows any of the three. No engagement letter, no fee, no confidentiality agreement signed by him, and no further exchange with him on this subject appears in this release15
Was anything sold?Not established, and the outside record points the other way. No price appears anywhere in these documents. Four years and nine months later a trade report stated that the house remained “under the control of Asim Abdullah, a San-Francisco based tech entrepreneur, and Aimz, his investment vehicle”16
The currency questionThe two licensee tables are headed “(Eur)” on their face1718. The profit-and-loss sheets carry no currency label and no unit label at all, and their figures are roughly a thousandth of the labelled tables’, so they appear to be thousands of euros; the sheets do not say so and this page does not supply it1920

Page limits. Nothing here is a finding of wrongdoing. A couture house looking for investment is ordinary commerce, and no person named on this page is alleged to have done anything wrong by appearing in these documents. Telephone, fax, mobile and email values printed or labelled on the documents cited here are not reproduced, and neither is any street address of a private individual. This page supplies no recipient and no sender from context for any of these messages, and makes no statement about how any producing party treated any field. Two classes of document reached by the same surname search were deliberately not opened, and a third was read only for its message stamps; all three are listed under What the surname reaches that this page is not about. Every quotation on this page was read on the page image of the sheet cited for it, at the resolution stated in its footnote, except where the footnote says otherwise.

In the third week of January 2011 the owner of the Paris couture house Emanuel Ungaro was looking for money. A friend introduced him to Jeffrey Epstein in two sentences of e-mail; a meeting was arranged in Paris for half past eight; and on the Sunday Epstein read the confidential book the house’s advisers had prepared and told its owner it was worthless as drafted. The book is physically in this release, all thirty-four sheets of it, with the licensee revenues, the royalty schedule, the operating expenses and the loss. What is not in this release, and what the outside record does not supply either, is any sale.

What the document actually is, and what it is not

The commission that produced this page described the episode as an attempted sale of the house. The memorandum does not support that description, and the correction matters because it is the difference between a transaction and an approach.

Sheet 19 of EFTA00521863 is headed “IMPORTANT LEGAL INFORMATION (1)”. Its second paragraph reads:

This draft Memorandum has been prepared and is being circulated on an introductory basis by Emanuel Ungaro (the “Company”). It is being delivered to a limited number of parties, who have already expressed a preliminary interest, in order to ascertain their interest in acquiring part of the share capital of the Company

and continues: “It does not constitute an offer or invitation to purchase any shares, and should in no way be perceived as investment advice, recommendations or services, and can not be construed as such offer, invitation, advice, recommendation or service in any jurisdiction where it would be illegal, or to the benefit of any person or entity which could not lawfully be the beneficiary of such offer…”4

Three things follow, and each is a separate fact. The book is a draft, by its own description. What it solicits is interest in part of the share capital, not the sale of the company. And it disclaims being an offer at all. One sheet later the same section says so again from the other end: “The issue of this Memorandum shall not be taken as any form of commitment on the part of Emanuel Ungaro or/and its shareholders to proceed with any transaction.”6

This is the shape of document that the desk’s own rulings warn about: an undertaking is not a completed act, and this is weaker than an undertaking. It is an invitation to express interest, hedged on its face against being read as anything more.

Field state. No sheet of the thirty-four carries a signature rule, an acceptance block, a date line or a price. There is nothing on it that anybody could execute, and nothing in the surrounding correspondence that anybody did. The cover is a pair of runway photographs above the house’s wordmark, with “INFORMATION MEMORANDUM” at the lower left and “DOSSIN ADVISORY” at the lower right, and it carries neither a date nor the name of a recipient.15 The only dates attaching to the book come from outside it: a covering e-mail of 19 January 2011 and the file names of its two attachments.10

The five questions a transaction page has to answer

What was sold. Nothing, so far as this release shows. What was offered for discussion was an unspecified minority of the share capital of Emanuel Ungaro.4

By whom. By the Company, whose sole owner since 2005 was Asim Abdullah, through Dossin Advisory.84

To whom. The memorandum names no addressee on its face.1 It reached Epstein on 23 January 2011. This page does not supply a forwarding party.1121

For what. Not established. No price, no valuation, no range and no multiple appears anywhere in the memorandum or in the correspondence read for this page. The memorandum sets out revenues, royalties, costs and assets, and stops there.1718192022

When. The book was reported complete on 18 January 2011; the covering e-mail attaching its two halves is quoted as of 19 January 2011; the approach to Epstein runs from 21 to 23 January 2011.2310914

How it reached him: the introduction, 21 and 22 January 2011

EFTA01799052 is two sheets and carries the whole introduction as a quoted chain. At the foot of it, timed in a quoted header to 7:11 p.m. on 21 January 2011, is the pitch:

> A new friend, of mine sold his Company for a lot of $s in the internet 1.0 boom and then bought the fashion house Ungaro in Paris > He has spent 50 million euro cleaning it up and is finally there > > It might be an opportunity if you had the right guy to run it > > He clearly is not ,as he will tell you > > He asked me to help him > > If you have the time (15 minuets) he is in Paris until Wednesday

The misspelling “minuets” and the comma placement are the sheet’s own.9 The sender is David J Mitchell, who has a page on this wiki; the fifty million euro figure is his, not the memorandum’s, and the memorandum nowhere states a figure for what the owner had put in.9

Epstein’s answer, quoted in the same chain at 8:05 p.m. the same evening, was six words long: “Yes karl lagerfelds assistant or mcqueens addy”.9 He was asked whether he was interested in a couture house and he replied with the names of two designers’ people. The following afternoon Mitchell wrote “When and where should he meet you in Paris.His name is Asim Abdullah”, Epstein answered with an arrangement, and Mitchell closed the exchange with “Will do”.9

The timings on this sheet are an interval, and here is the arithmetic. Epstein’s reply carries an explicit zone: “Date: Sat, 22 Jan 2011 15:03:42 +0100”, which is 14:03:42 UTC. Mitchell’s “Will do” carries the producing party’s own header, “Sent: Saturday, January 22, 2011 2:19 PM”, which on this document family is UTC (see The clocks below), so 14:19:00 UTC. 14:19:00 minus 14:03:42 is +00:15:18. The sign control is that a reply must follow the message it answers, and it does.924

The meeting that was arranged, 22 January 2011

At 15:32:32 UTC on 22 January, an hour and twenty-nine minutes after telling Mitchell where to send him, Epstein wrote to a correspondent this page does not identify: “Great the owner of ungaro is coming at 830”.13

That interval is zoned against zoned, and here is the subtraction. Epstein’s arrangement carries “Date: Sat, 22 Jan 2011 15:03:42 +0100”, and Paris stood at Central European Time, UTC+1, on 22 January 2011, European summer time not having begun until 27 March: that instant is 14:03:42 UTC. The 8:30 note carries “Date: Sat, 22 Jan 2011 15:32:32 +0000”, which is 15:32:32 UTC. 15:32:32 minus 14:03:42 is +01:28:50. The sign control is that the later act must follow the earlier, and it does. ⚠ Subtracting the two printed clocks without converting either gives twenty-nine minutes, and this page carried that figure until it was corrected: the error is exactly the Paris offset, and the same stamp is converted correctly four lines above.91324

That sentence stands in six separate documents in this release and says the same thing in all six.25 It is an arrangement and not an occurrence. No document read for this page records the meeting taking place, who was there, or what was said. Nothing here should be read as establishing that the two men met.

The designer, late on 22 January 2011

Some seven and a half hours later Epstein sent a one-line enquiry to a third party, again not identified here: “giles deacon„ now with ungaro. —reputation.. does he have any talent?”12 ⚠ The doubled comma is the extraction’s rendering and this page does not say it is the paper’s: the sheet was not rendered for this page, for the reason given in the next paragraph, and a comma read as a doubled mark is one of the commonest artefacts in this release.

Giles Deacon had been the house’s creative director since the summer of 2010, and the memorandum devotes two sheets to him, including a slide of press praise quoting a Los Angeles Times critic hoping “he can inject some cool into Ungaro too” and a New York Times critic calling him “a 40-year-old designer with a couture sensibility who has developed his own house and could make a serious job of resuscitating the Emanuel Ungaro heritage”.2627 Nothing in Epstein’s question is an allegation about Deacon, and nothing on this page should be read as one; it is a buyer’s enquiry about a brand’s principal creative asset. No statement by Deacon about this correspondence was found in the material searched for this page, and one web search was run on the sale of the house without any search being made for a response from him.

⚠ The enquiry was answered, and this page does not reproduce the answer. It stands at the foot of a longer exchange on the sheet cited here, running on to 26 January 2011. The person who answered is a named private individual whose status this page has not established and does not attempt to establish; nothing from that exchange is quoted, counted or characterised here, and no search was run on that person.12 ⚠ An earlier version of this page said the record contained no answer. That was false, and it was refuted by the document the sentence itself cited.

This message is the single most replicated item in the episode: twenty-six separate documents in this release carry it. Two of those twenty-six are reached by the house’s name but not by a search for the question itself, because the extraction of those two copies damages a word inside the sentence; the damage class is quoted-printable encoding and the damaged token is not reproduced here. ⚠ Twenty-six documents are not twenty-six copies of one message. Of the twenty-four reached by the question’s own words, twenty-one carry more than one message and are chains in which the question is quoted; the remaining three are the single message on its own.28

The forwards, 23 January 2011

Five documents record the book and its underlying files being pushed to Epstein across forty-five minutes on the Sunday afternoon, from 17:45 to 18:30 UTC.

At 17:44 UTC Abdullah wrote to the intermediary “Kindly forward to Jeffrey…its in 2 parts”, and that instruction was acknowledged a minute later with “Great. Thank you”.11 Quoted beneath it in the same chain is the advisory firm’s own covering e-mail of 19 January, addressed to the owner and to the company’s General Manager:

Dear Asim and Marie,

I am resending this in 2 different files, in case my previous mail was not delivered because it was too big.

Please find attached the last version of the Information Memorandum.

Asim, could you please confirm that you are happy for us to send it to potential investors?

The attachment is named in the chain as “Info Memo - 19th jan ‘11 - Part 1.pdf”.10

⭐ That last line is the sharpest thing in the file. On 19 January the house’s adviser was still asking the owner for permission to send the book to potential investors. Four days later the same book was in Epstein’s inbox, and the person who sent it on was the owner himself, whose permission the adviser had asked for. The record does not contain the owner’s answer to that question, so what the sequence raises is a question rather than a finding: was the book circulated with the adviser’s clearance, or ahead of it? This page does not answer it, and the two dated facts are the whole of what the documents show.

Parts 1 and 2 were sent on to Epstein at 17:49 and 17:52 UTC. Both figures come from Sent: headers, which are transmission times and not delivery times, and no document read for this page records when anything arrived.2129 A further forward at 18:03:58 UTC carried an offer of travel: “Also ▮.I am in paris till fri but can travel to ny on my way to sf if Jeffrey wants me to meet someone there…”, where ▮ marks a producer’s mark on the sheet, read at the page image at 600 dpi; the page’s earlier rendering of that passage printed an ellipsis in its place, which made a withheld passage look like the writer’s own punctuation.30 That too is an offer and not an occurrence; no document read for this page shows such a trip or such a meeting. At 18:30 UTC Abdullah forwarded the underlying workbooks, three of them, named in the header as a personnel-cost file, a three-year cash-flow file and a three-year plan.31

What Epstein wrote back

At 18:38:50 UTC on 23 January 2011 Abdullah wrote: “Have just forwarded the detailed financials..some additional notes…”, and set out four, which are the only terms of the house’s capital structure stated anywhere in this episode:

  1. On the Ferragamo loan…3m euros if paid by June 30, 2011 or 1m euros per year for 4 years starting year end 2011

  2. Perfume licensed to the Ferragamos till 2015…buy back deal would have to be negotiated with them as there is no pre-set price

  3. U by Ungaro license buy back option comes available June 30, 2012 for 350M yen…throws off incremental 2 m euros in lic revenue per year in min guaranteed revenues

  4. Assets include the key money on the boutique (8-10m euros), the ip and a couture dress collection valued at 3m euros…

The currencies are stated on the sheet: euros in notes 1, 3 and 4, yen in note 3.14 The dates in that note are field-typed and none of them is a closing date. “June 30, 2011” is a loan repayment deadline; “2015” is a licence term end; “June 30, 2012” is the date an option becomes exercisable. No date in this episode is the date of any transaction, because there is no transaction.

Quoted immediately beneath it, under a header whose clock is dealt with in the next section, is Epstein’s message in full:

A highly promoter type presentation.. I could n;t show it to anyone. i would beed working capital needs. licencing deals. what percentages on what items, note details to ferragamo, and perfume buyback. This presentation will not do you any good unless an arab russian or chinese has a girlfriend that wants to be in the biz

Every irregularity in that passage is on the paper, and the release carries that finding six ways over. Read at 600 dpi, the sheet prints “n;t”, not “not”; “beed”, not “need”; and “licencing” with a c. These are the writer’s own keystrokes, not damage, and they are reproduced rather than repaired.14

⭐ The image is not the only witness, and this page understated its own evidence until the point was put to it. The release holds six copies of this message. Five of the six print “n;t” in the machine-readable text layer, where any reader can check it without rendering anything at all; the sixth is the copy cited for the transcription above, whose extraction gives a different character in that position. Two of the six carry the whole transcribed passage, “n;t” and “beed” and “licencing” together, verbatim in the text layer. So the commission’s “I could n[o]t show it to anyone” is corrected not by one image reading but by five text layers and a 600 dpi render. The three extractions that differ do so only in the second and third tokens; the damage classes are character substitution and quoted-printable encoding, and the damaged tokens are not reproduced here.32

The message is three acts in four lines. It rejects the book as a selling document; it lists the five things a usable one would contain, four of which the owner then supplied within the hour; and it closes with a sentence about who the buyer for such a brand would be.

The eight-hour trap, and what the sequence really is

Read only in the copies that quote it, Epstein’s critique appears not to fit. The quoted header above it says 10:27:10 on Sunday 23 January, while every delivery of the memorandum that this release contains is stamped between 17:44 and 18:30 UTC the same day. Taken at face value he would have condemned a presentation some seven hours before any copy of it reached him.

It is an artefact of a third clock, and the release settles it twice over. Epstein’s message also survives in this release as he sent it, in two copies that are not replies and carry no quotation, and on those the producing party’s own header reads “Sun 1/23/2011 6:27:10 PM” and “Sunday, January 23, 2011 6:27 PM”.3334 That is 18:27:10 UTC.

⭐ And one of those two copies carries a machine field rather than a human one, which fixes the instant independently of every header on this page. In its trailing block of mail-client metadata it prints date-received 1295807230, a Unix epoch value. Converted: 1295807230 is 18:27:10 UTC on 23 January 2011, exactly, and the same instant renders as 10:27:10 under an offset of minus eight hours, exactly. A machine-written integer is the strongest evidence a zone claim can have here, and it agrees with the produced headers to the second.34

The subtraction is exact: 18:27:10 minus 10:27:10 is 8h00m00s, and the seconds agree to the digit. Eight hours behind Coordinated Universal Time on 23 January 2011 is US Pacific Standard Time, daylight saving in the United States not having begun until 13 March that year. ⚠ What the offset establishes is an offset of minus eight hours, not a place. Any locale standing at UTC−8 that day, or a mail client merely set to it wherever its owner happened to be, would print the same figure. That the clock is a Pacific one rests on the trade press description of the man who replied as San Francisco based, which this page cites and does not treat as more than it is.2416

The sign control is that a west-of-Greenwich local clock must read earlier than the same instant in UTC, and 10:27 is earlier than 18:27. A one-hour or five-hour offset would have put it in Paris or New York; neither reconciles, and only the same-message produced twin recovers the zone. No conversion of an unzoned quoted header is made anywhere else on this page.

With that resolved the week closes without a gap, and the pace is the finding:

UTC, 23 January 2011What happened
17:44:05 and 17:44:20The owner forwards parts 1 and 2 with “Kindly forward to Jeffrey…its in 2 parts”29
17:45”Great. Thank you”11
17:49Part 1 sent on to Epstein10
17:52Part 2 sent on to Epstein21
18:03:58A further forward, carrying the offer to travel30
18:27:10Epstein’s critique33
18:30The owner forwards three workbooks31
18:38:50The owner’s four notes: “Have just forwarded the detailed financials”35

Epstein read a thirty-four-sheet confidential memorandum and rejected it thirty-eight minutes after the first half of it was sent on to him, and the owner supplied four of the five things he had asked for eleven minutes and forty seconds after that. Both intervals are zoned against zoned and both signs are positive.36

One thing the resolution does not settle is whether the two men had already met. A meeting was arranged for the Saturday evening and this release does not show it happening, so whether Epstein came to the book cold on the Sunday afternoon or had been shown it the night before is not established here.

The clocks

Every stamp in this episode has been classified, and the producing party’s own Sent: header is UTC on this document family. That is not assumed; it is established on three exact same-message pairs, each pairing a produced header against a copy of the identical message carrying an explicit offset, matching to the second.24

MessageZoned copyProduced copyAgreement
Abdullah’s four notes, 23 Jan”Sun, 23 Jan 2011 18:38:50 +0000"35"Sun 1/23/2011 6:38:50 PM”14exact to the second
”the owner of ungaro is coming at 830”, 22 Jan”Sat, 22 Jan 2011 15:32:32 +0000"13"Sat 1/22/2011 3:32:32 PM”37exact to the second
the designer enquiry, 22 Jan”Sat, 22 Jan 2011 23:04:10 +0000"38"Sat 1/22/2011 11:04:10 PM”39exact to the second

⭐ And the third row carries a trap that this release sets for any reader. The same designer enquiry is stamped “Sun, 23 Jan 2011 00:04:10 +0100” in other copies.12 It is one message at one instant, and the release gives it two different dates: 22 January on the UTC copies and 23 January on the Paris copies, because 23:04:10 UTC rolls past midnight into Central European Time. 23:04:10 plus one hour is 00:04:10 on the following day, and the sign control is that adding a positive offset moves a clock forward. A page taking the date off either copy alone would be half wrong.

A fourth pair is the one that matters most, and it is set out in full in the section above: Epstein’s critique, quoted at “Sun Jan 23 10:27:10 2011” and produced at “Sun 1/23/2011 6:27:10 PM”, an exact eight-hour offset. The two stamps stand on different documents and each is cited for its own: the quoted 10:27:10 header is on the copies carrying the owner’s reply above it, and the produced 6:27:10 PM header is on the copy of the message as sent.143533

The unzoned stamps are these, and apart from the critique just discussed none of them is used here to fix an hour: “Sent : Tue Jan 18 09:07:16 2011” (the book reported complete), “Sent: Wed Jan 19 17:00:49 2011” and “Wed Jan 19 17:03:54 2011” (the adviser’s two covering e-mails, the second of which stands on the part 2 forward and not on either sheet cited for the first), “Jan 21, 2011 8:05 PM”, “Jan 21, 2011, at 7:11 PM” and “On Sat, Jan 22, 2011 at 2:53 PM” in the introduction chain, and “On Jan 22, 2011, at 4:31 PM” beneath the 8:30 note.23102129913 Two stamps inside the forwarding chain do carry a zone, “January 23, 2011 6:44:05 PM GMT+01:00” and “6:44:20 PM GMT+01:00”, which are 17:44:05 and 17:44:20 UTC.2129

⚠ The general lesson this page had to learn about itself: an unzoned quoted header does not merely lack an hour, it can silently carry a zone nobody in the conversation is standing in. The owner was writing from Paris that week and his client was stamping in California. The only thing that recovered it was a copy of the same message under a producing party’s own header, and the only reason that copy was found was a search run on the words of the quotation itself.

⚠ One pairing on these sheets will mislead a careless reader in the opposite direction. A message stamped 6:44 p.m. in a quoted Paris header is answered by a message the producing party stamps 5:45 p.m., so the answer appears on the face of the sheet to precede the thing it answers by an hour. It does not: 18:44 CET is 17:44 UTC and the reply is 17:45 UTC, one minute later.11

What the book contains

It is not marketing boilerplate. Twenty-four of its thirty-four sheets carry the house’s name, and about a third of it is financial.

The licensee table on sheet 7 of EFTA00521863, headed “(Eur)”, gives total wholesale sales by licensees of 53,116,358 for 2009 against a 2010 estimate of 73,150,000, broken down by region and by line, with Japan shown separately at 15,815,137 for 2009.17 The royalty table on the next sheet, also headed “(Eur)”, gives a 2009 grand total of 3,177,114 against a 2010 estimate of 3,411,150.18

The profit-and-loss sheets are where the house’s position shows. Sheet 17 gives total net revenues of 8,475 for 2009 falling to 6,471 for 2010, against projections of 14,352, 22,625 and 30,482 for 2011, 2012 and 2013.19 Sheet 18 gives total costs of 14,606 for 2009 and an EBITDA of negative 11,046, at minus 130 per cent of net revenues, improving to minus 5,323 in 2010 and projected to turn positive at 2,043 in 2012.20 Neither sheet states a currency or a unit. Set against the labelled tables these figures appear to be thousands of euros, which would make the 2009 loss about eleven million euros; the sheets do not say so, and this page does not supply what they omit.

The memorandum does not hide the loss. Its own risk section says in terms: “Emanuel Ungaro has been reporting losses in recent years. Whilst results have been improving and management is looking for the Company to reduce losses, and then to be profitable in future years, it is worth mentioning that, should these losses continue, they would have to be financed.”40

The assets sheet claims a flagship store of “some 360 sqm” on Avenue Montaigne with a key money value “estimated at Eur9.5m about 3 years ago”, about 730 square metres of offices and showrooms above it, ownership of the DIVA brand name in perfume and cosmetics, and a tax-loss carry-forward the sheet prints as “c. Eur80mn”. It ends: “The company is virtually clear of bank debt.”22

⚠ The key money figure and the owner’s figure do not agree. The memorandum, undated on its face but attached to e-mails of 19 January, says Eur9.5m as estimated three years earlier. Abdullah’s note of 23 January says “8-10m euros”. The two are not reconcilable into one number and both are reproduced here as they stand.2214

The restructuring sheet is candid about what a buyer would be taking on: “Key management positions (Chief Executive Officer and Chief Financial Officer) are available to be filled by the future shareholder/operator”, with “Only c.25 staff members”.41 A 2010 organisation chart on the following sheet shows the owner as President and CEO above a General Manager, over studios, atelier, sales, communications and finance departments whose stated headcounts total thirty-nine, and is marked CONFIDENTIAL. Only two people are named on the chart.42

The book closes on a single line: “ONE OF THE MOST RECOGNIZED COUTURE HOUSE”, the missing plural being the sheet’s own.3

What happened next

Nothing in this release. No further exchange between Epstein and the owner on this subject was found in the material searched for this page; no engagement, fee, mandate, confidentiality agreement, term sheet or further draft appears; and the episode stops on 23 January 2011.15

Outside the release, and searched for this page on 18 October 2026, the public record does not record a sale either. A 2005 trade report, fetched and read in full, states that “Gruppo Ferragamo signed an agreement to sell 100 percent of the company to San Francisco-based high-tech entre pr eneur Asim Abdullah”, that terms were not disclosed, and that the Ferragamos “will continue to hold the Ungaro licenses for leather goods and fragrances”, which corroborates both the memorandum’s ownership slide and Abdullah’s note about the perfume licence.43 A 2015 trade report, also fetched and read in full, states that the house was then “under the control of Asim Abdullah, a San-Francisco based tech entrepreneur, and Aimz, his investment vehicle”.16 So four years and nine months after the memorandum reached Epstein, the owner it was issued for still owned the company. The holding vehicle named in that report, Aimz, returns zero documents in this release.

What the outside record does report between those two points is a licence rather than a sale: in September 2012 the Italian group Aeffe took over production and distribution of the house’s womenswear, an arrangement the parties did not renew in 2015. This page has not opened a contemporaneous report of that 2012 agreement and does not state its terms; a fetch of one such report returned an HTTP 403 interstitial carrying no article text, and the claim is not relied on here.44

Who is named here, and who is not

Three people are named: Asim Abdullah, who owned the company and wrote the messages; David J Mitchell, who made the introduction and whose own page on this wiki does not yet carry this episode; and Giles Deacon, the house’s creative director, whom the memorandum names and praises across two sheets.

Two more are referred to by office in this page’s own prose: the company’s General Manager and the principal of Dossin Advisory. ⚠ Both names nevertheless stand on this page, inside quotations of documents it cites and links: the principal’s in the legal-notice passages quoted below, the General Manager’s in the adviser’s salutation and in a quoted address line. This page therefore does not withhold either name, and an earlier version of it said three times that it did. Each is a professional acting in a professional capacity in an ordinary commercial transaction, and each name is a narrow search key, reaching two or three documents in this release, every one of which this page already cites and links. Whether a page should name officers and advisers acting in their office, in its own voice as well as in quotation, is an open editorial question that this page does not settle. The advisory firm is named throughout, because its name is printed on the footer of almost every sheet of a document this page cites and links.

This page withholds the recipients of two of Epstein’s messages entirely, and the forwarding party in the memorandum chain entirely. In each case the standing rule is to withhold rather than to choose, and this page supplies no sender and no recipient from context. It makes no statement about how any producing party treated any field, on any copy, because a claim of that kind cannot rest on a text layer and the sheets concerned were not rendered for this page.

What the surname reaches that this page is not about

A search of the whole corpus for the token Ungaro returns 67 distinct documents across 91 sheets. Forty-one of the sixty-seven concern this episode and twenty-six do not, and of those forty-one, thirty-eight are duplicate copies of just four short messages, leaving the memorandum’s two halves and the introduction chain. The rest are a different person, a different decade or a different industry relationship, and the breakdown is given here so that the figure is not mistaken for a measure of the record on this subject.

ClassDocumentsSheetsWhat it is
Judge Ursula Ungaro, S.D. Fla.1315Case captions such as 15-24585-CV-UNGARO and 11-20216-CIV-UNGARO, an expert’s curriculum vitae, trial-scheduling correspondence and a law firm’s newsletter. A United States District Judge who shares the surname. Nothing to do with the house
The information memorandum224EFTA00521855 and EFTA00521863
The 23 January exchange44Four copies of the message pair, each carrying the owner’s reply above Epstein’s message quoted beneath it
The book-completed forwards22Two forwards of one internal message
The Mitchell introduction11EFTA01799052
”the owner of ungaro is coming at 830”66Six copies of one message
The designer enquiry2626Twenty-six documents carrying one message. Not twenty-six copies of it: of the twenty-four reached by the question’s own words, twenty-one are chains carrying more than one message, and the twenty-six documents stand on forty-nine sheets in all, of which twenty-six carry the house’s name
February 2011 private correspondence55Five copies of one personal message to Epstein that mentions a casting at the house’s showroom in passing. Not opened, not cited and not characterised here, and no search was made on its writer
2017 correspondence66A thread of 7 and 8 October 2017, six years after this episode, in which the house is named. Not six copies of one message: counting stamps alone, without reading any body, the six documents carry eight distinct message stamps and at least five distinct messages. Not cited and not characterised here: nothing was read from these six documents but their message stamps, counted to establish that they are a thread and not six copies, and no search was made on any person in them
A modelling portfolio sheet11A sheet listing the house among a person’s advertising credits. Not opened, not cited and not characterised here; it carries personal particulars of a named individual
An advertising agency’s client list11The house named among a Paris agency’s campaign clients
Total6791

Six further documents belong to this episode and carry no occurrence of the house’s name at all, so the token search misses them. Four are covering e-mails that name only the people: EFTA02034543, EFTA02320753, EFTA02320837 and EFTA02375967, reached by a search for the owner’s name, which returns thirteen documents of which those four lie outside the sixty-seven. Two are Epstein’s critique as he sent it, EFTA01834414 and EFTA02534817, which quote nothing and therefore never mention the house: they were reached only by searching the words of the quotation itself, and they are the two documents that resolve the clock. So the episode stands in 47 documents, not 41, and the union of the three arms is 73: 41 episode documents inside the 67, 26 non-episode documents inside the 67, and 6 episode documents outside it.

⚠ That is the single most consequential measurement on this page, and the page was written once without it. A floor built on the house’s own name would have missed the two sheets that date the central message, and the page would have carried a false finding about a gap in the record. A name is not a probe for a message that does not name anything.

The forms searched, each as its own arm, in distinct documents. Ungaro 67; Emanuel Ungaro 4; Emanuel Ungaro SAS 2; Ungaro SAS 2; Ungaro Paris 2; U by Ungaro 5; Asim Abdullah 13; Dossin Advisory 2; Dossin 5, of which one is not the firm; Ferragamo 99; Aeffe 2, neither about this house; Aimz 0. Negative controls: Ungarozzz 0 and Aimzzz 0. Two spellings that might have been expected return nothing and are worth recording as clean negatives: Emmanuel Ungaro, with two Ms, returns 0, although that is the commoner spelling of the forename, and Ungaro SA returns 0 while the company’s actual legal form, Ungaro SAS, returns 2. Ungar, a different surname, returns 61 documents and shares not one with the 67.

Three single-edit variants of the house’s name return one document each. Two of them, on sheets 13 and 15 of EFTA00521863, are the house’s own wordmark defeating the extraction on slides of a document already cited here, and they reach no new document; one is unrelated character noise in a government form. The wordmark is set in a rounded lowercase display face and the extraction renders it differently on almost every sheet of the memorandum, which is why ten of the book’s thirty-four sheets are not reachable by its own name.45

Notes on the sources

Footnotes

  1. Information memorandum, sheet 1 of 8 of the run stamped EFTA00521855 to EFTA00521862: a landscape slide carrying two runway photographs above the wordmark “emanuel ungaro” in rounded lowercase magenta type, with “P A R I S” beneath it, “INFORMATION MEMORANDUM” at the lower left and “DOSSIN ADVISORY” at the lower right. There is no date anywhere on the sheet, no addressee, no recipient line, no draft stamp and no confidentiality legend. The extraction renders the two photographs as the token “43%-J”, which is not text on the sheet. Read from the page image at 400 dpi. https://epstein-data.com/EFTA00521855 p. 1. ↩ ↩2 ↩3 ↩4

  2. Same document, sheet 2 of 8, headed “TABLE OF CONTENT”, a sixteen-item list. Counted on the page image: sixteen bullets and sixteen page numbers, the page numbers set as a right-aligned second column on the same line as each item, not on the line below it. (The extraction reflows that second column onto separate lines, and an earlier version of this footnote took the layout from the extraction and gave the count as seventeen.) The first item is “Legal Notice” against page 3 and the last is “Important Legal Notice” against page 27; others include “A Licensing Model Offering Full Flexibility Going Forward” against page 15, “Restructuring Program Under Way” against page 22, “Attractive Assets and A Particularly Clean Balance Sheet” against page 24 and “Financials” against page 25. The memorandum’s own internal page numbering therefore runs to at least 27 while the release holds 34 sheets under two identifiers; this page cites by release identifier and sheet, never by the book’s internal numbering. Read from the page image at 400 dpi; the item count and the two-column layout were settled there and not at the text layer. https://epstein-data.com/EFTA00521855 p. 2. ↩

  3. Same document, sheet 26 and the last of the run, carrying the wordmark above the single line “ONE OF THE MOST RECOGNIZED COUTURE HOUSE” and the footer “DOSSIN ADVISORY”. The missing plural on “HOUSE” is the sheet’s own. Read from the page image at 400 dpi, where the line stands as quoted, missing plural included. https://epstein-data.com/EFTA00521863 p. 26. ↩ ↩2

  4. Information memorandum, sheet 19 of the 26-sheet run stamped EFTA00521863 to EFTA00521888, headed “IMPORTANT LEGAL INFORMATION (1)”. The quoted English paragraph begins “This draft Memorandum has been prepared and is being circulated on an introductory basis by Emanuel Ungaro (the “Company”)” and continues through “It does not constitute an offer or invitation to purchase any shares, and should in no way be perceived as investment advice, recommendations or services, and can not be construed as such offer, invitation, advice, recommendation or service in any jurisdiction where it would be illegal…”, the sentence running on past the point quoted. The sheet’s third paragraph names the advisory firm and its principal as persons giving no warranty. The sheet opens with a French-language paragraph which is not quoted here and from which nothing on this page is taken. Read from the page image at 400 dpi. https://epstein-data.com/EFTA00521863 p. 19. ↩ ↩2 ↩3 ↩4 ↩5 ↩6

  5. A statement about all 34 sheets of the two runs, made after reading every sheet at the text layer and six at the page image. No sheet carries a signature rule, an initial block, an acceptance block, a date line, a price, a valuation or a transaction date. There is no instrument in this episode capable of execution: the memorandum is a presentation and the remaining documents are e-mails. Sheets 1, 4, 7 and 19 of EFTA00521863 and sheets 1 and 2 of EFTA00521855 were read at the page image; the other 28 at the text layer only. https://epstein-data.com/EFTA00521863 p. 1. ↩ ↩2

  6. Same document, sheet 20, headed “IMPORTANT LEGAL INFORMATION (2)”: “Emanuel Ungaro, Jacques-Franck Dossin or/and Dossin Advisory and their Related Entities or Persons undertake no obligation to provide the recipients of this Memorandum with access to any additional information or to correct any inaccuracies herein which may become apparent, and they reserve the right, without advance notice, to change the procedure related to this Investment Opportunity or to terminate negotiations at any time prior to agreeing any offer. The issue of this Memorandum shall not be taken as any form of commitment on the part of Emanuel Ungaro or/and its shareholders to proceed with any transaction.” Read from the page image at 400 dpi, where the passage stands as quoted. https://epstein-data.com/EFTA00521863 p. 20. ↩ ↩2

  7. Same document, sheet 25, headed “IMPORTANT LEGAL INFORMATION (7)”: “Jacques-Franck Dossin and/or Dossin Advisory and their Related Entities or Persons are acting only for the Company in connection with this Investment Opportunity and is not acting for or advising any other person.” This is the sheet establishing that the adviser acted for the house and for nobody on the other side. Read from the page image at 400 dpi, where the passage stands as quoted, the singular verb after the plural subject included. https://epstein-data.com/EFTA00521863 p. 25. ↩

  8. Same document, sheet 5 of 8, headed “History and Essence of the Brand”. Under “Ownership has been as follows:” the sheet gives three bullets: “‘65-‘95 : Brand owned and managed by Mr Ungaro”; “‘95-‘05 : Ownership transferred to the Ferragamo family who operated a shift in strategy to reintegrate some business lines internally and to enlarge the directly-operated store (DOS) network”; “Since ‘05 : Ownership transferred to Mr Asim Abdullah”. The wordmark at the head of this sheet is rendered by the extraction as “emanuel ung€ro”; at the image it is the house’s ordinary wordmark. Read from the page image at 400 dpi, where all three bullets stand as quoted. https://epstein-data.com/EFTA00521855 p. 5. ↩ ↩2

  9. E-mail chain, EFTA01799052, sheet 1 of 2, stamped EFTA_R1_00137430. The produced header reads “From: DAVID J MITCHELL”, “Sent: Saturday, January 22, 2011 2:19 PM”, “To: Jeffrey Epstein”, “Subject: Re: Good to see you”, and the body is “Will do” over the sign-off “DAVID MITCHELL”. Quoted beneath: a “From:” line naming Jeffrey Epstein and a “Date:” line reading “Sat, 22 Jan 2011 15:03:42 +0100”, whose one-line body carries contact particulars and is not reproduced here. Below that: “On Sat, Jan 22, 2011 at 2:53 PM, DAVID J MITCHELL” wrote “When and where should he meet you in Paris.His name is Asim Abdullah”; then “Sent: Jan 21, 2011 8:05 PM” over “Yes karl lagerfelds assistant or mcqueens addy” and “Sorry for all the typos .Sent from my iPhone”; then “On Jan 21, 2011, at 7:11 PM, “DAVID J MITCHELL"" over the six quoted lines reproduced in the body of this page. The extraction renders the middle initial of one of those attributions as the digit 1; the image reads J. Read from the page image at 400 dpi. https://epstein-data.com/EFTA01799052 p. 1. ↩ ↩2 ↩3 ↩4 ↩5 ↩6 ↩7 ↩8 ↩9

  10. E-mail forward, EFTA02375967, a single sheet stamped EFTA_R1_01384885: produced header “Sent: Sunday, January 23, 2011 5:49 PM”, “To: Jeffrey Epstein”, “Subject: Fwd: Information Memorandum - part 1”, “Attachments: Info Memo - 19th jan ‘11 - Part 1.pdf; Untitled attachment 00046.htm”. Beneath: a “From:” line naming Asim Abdullah, a “Date:” line reading “January 23, 2011 6:44:05 PM GMT+01:00”, a “Subject:” line reading “Fw: Information Memorandum - part 1”, and the body “Kindly forward to Jeffrey…its in 2 parts”. Beneath that, the adviser’s own e-mail of “Sent: Wed Jan 19 17:00:49 2011”, addressed “To: Asim Abdullah; Mfournier”, whose body is quoted in full in the text of this page and whose sign-off is the adviser’s name. No forwarding party is supplied here, and this page makes no statement about the state of that field on this or any copy: it was not read at the image. Text layer only. https://epstein-data.com/EFTA02375967 p. 1. ↩ ↩2 ↩3 ↩4 ↩5 ↩6

  11. E-mail, EFTA02320753, a single sheet stamped EFTA_R1_01230362: produced header “Sent: Sunday, January 23, 2011 5:45 PM”, “To: Asim Abdullah”, “Subject: Re: Information Memorandum - part 1”, body “Great. Thank you”. Beneath: “On Jan 23, 2011, at 6:44 PM, Asim Abdullah” wrote “Kindly forward to Jeffrey…its in 2 parts”, above the same 19 January adviser e-mail, here signed in a form the extraction damages. This sheet is the one on which a quoted 6:44 p.m. Paris stamp sits above a produced 5:45 p.m. UTC stamp. No sender is supplied here, and no statement is made about the state of that field: it was not read at the image. Text layer only. https://epstein-data.com/EFTA02320753 p. 1. ↩ ↩2 ↩3 ↩4 ↩5

  12. E-mail chain, EFTA00648317, two sheets: “From: Jeffrey Epstein”, “Date: Sun, 23 Jan 2011 00:04:10 +0100”, body “giles deacon„ now with ungaro. —reputation.. does he have any talent?” above the standard footer. The sender’s address is a contact-class value and is not reproduced here. ⚠ The enquiry is the earliest message on the sheet, not the only one. Above it, later in time, the chain carries further messages under “Date:” headers of 23, 25 and 26 January 2011, among them an answer from the person the enquiry was addressed to. That person is a named private individual whose status this page has not established; the page does not name, quote, count or characterise anything from those messages, no search was run on that person, and the sheet was not rendered. No recipient is supplied here, and this page makes no statement about the state of that field on any copy. The doubled comma after the surname stands in the extraction and is not asserted here to be on the paper. Text layer only, and the structure above was read from the chain’s own “Date:” and “On … wrote” lines without reading any body. https://epstein-data.com/EFTA00648317 p. 1. ↩ ↩2 ↩3 ↩4

  13. E-mail, EFTA00679596, a single sheet: “From: jeffrey epstein”, the sender’s address being a contact-class value not reproduced here, “Subject: Re: No subject”, “Date: Sat, 22 Jan 2011 15:32:32 +0000”, body “Great the owner of ungaro is coming at 830” above “Sony for all the typos .Sent from my iPhone”, above a quoted line timed “On Jan 22, 2011, at 4:31 PM”. The recipient field is not reproduced here and no recipient is supplied. “Sony” for “Sorry” is the extraction’s. Text layer only. https://epstein-data.com/EFTA00679596 p. 1. ↩ ↩2 ↩3 ↩4 ↩5

  14. E-mail, EFTA02002690, a single sheet stamped EFTA_R1_00499984. Produced header: “From: Asim Abdullah”, “Sent: Sun 1/23/2011 6:38:50 PM”, “Subject: Re:”. Body: “Have just forwarded the detailed financials..some additional notes…” and the four numbered notes quoted in full in the text of this page. Quoted beneath: a “From:” line naming Jeffrey Epstein, a “To:” line naming Asim Abdullah, a “Sent:” line reading “Sun Jan 23 10:27:10 2011”, and an empty “Subject:” line, above the reply quoted in full in the text of this page. Read at 600 dpi: the sheet prints “I could n;t show it to anyone”, “i would beed working capital needs” and “licencing deals”, and all three stand as the writer typed them. The text layer of this copy renders “n;t” as “n3”, which the image refutes. The recipient field of this message carries a contact-class value which is not reproduced here and about which this page makes no statement. Read from the page image at 600 dpi. https://epstein-data.com/EFTA02002690 p. 1. ↩ ↩2 ↩3 ↩4 ↩5 ↩6 ↩7

  15. An absence scoped to what was searched, not a claim about the release. Column-scoped full-text queries were run over the whole corpus for the house’s name in twelve forms, for the owner’s name in two, for the advisory firm in two and for its principal in one, and every document returned was partitioned into the classes tabulated above. No engagement letter, retainer, fee note, invoice, term sheet, confidentiality agreement, share purchase agreement or subsequent draft naming both Epstein and this house was among them, and no exchange between Epstein and the owner later than 23 January 2011 on this subject was among them. The forms searched and their counts are printed in the body of this page so that the search can be repeated. https://epstein-data.com/EFTA02002690 p. 1. ↩ ↩2

  16. FashionUnited, “Ungaro and Aeffe part ways”, dated 2015 on the page, at https://fashionunited.uk/news/business/ungaro-and-aeffe-part-ways/2015100617899 , fetched 18 October 2026, HTTP 200, 66,384 bytes, content-type text/html. The article states: “Emanuel Ungaro is under the control of Asim Abdullah, a San-Francisco based tech entrepreneur, and Aimz, his investment vehicle.” It also reports that Emanuel Ungaro and Aeffe SpA “have decided not to renew the license they signed three years ago”, attributing the report to WWD. ↩ ↩2 ↩3

  17. Information memorandum, EFTA00521863 sheet 7, headed “A Licensing Model Offering Full Flexibility Going Forward” above “Summary of Net Wholesale Turnover by Licensees”, the table headed “(Eur)” with columns “2,009”, “2010 E”, “2011 E”, “2012 E”, “2013 E”. Row “Total Wholesale Sales by Licensees”: 53,116,358 / 73,150,000 / 90,560,000 / 104,100,000 / 102,400,000. Row “Total Sales Excl. Japan” for 2009: 37,301,221. Row “Total Japan” for 2009: 15,815,137, and 37,301,221 plus 15,815,137 is 53,116,358 exactly. The extraction presents this table’s columns as separate vertical runs, so individual cells are not paired by position here and only the labelled total rows are reported. ⚠ Read from the page image at 600 dpi, and the row label is the reason. This sheet’s extraction renders “Excl.” as “Exd.”, welding the c and the l into a single letter; the image prints “Total Sales Excl. Japan”, clean. An earlier reading of this page treated the extraction’s form as the sheet’s and would have published it as the row’s name. The four estimate-column labels each carry a space before the E on the sheet, which an earlier version of this footnote dropped in three of the four. https://epstein-data.com/EFTA00521863 p. 7. ↩ ↩2 ↩3

  18. Same document, sheet 8, headed “Economic Size of the Brand much Larger than Current Company’s Revenue” above “Summary of Royalties Revenue”, the table headed “(Eur)”. Row “Grand TOTAL”: 3,177,114 / 3,411,150 / 4,486,750 / 6,261,000 / 7,607,000 across 2009 and the four estimate years. Same column-run caveat as the preceding sheet. Read from the page image at 400 dpi, where the heading, the “(Eur)” label and every figure in the quoted row stand as given; this footnote quotes none of the sheet’s column labels. https://epstein-data.com/EFTA00521863 p. 8. ↩ ↩2 ↩3

  19. Same document, sheet 17, headed “SALES AND GROSS PROFIT”, columns “2009”, “2010 F”, “2011 E”, “2012 E”, “2013 E”. Row “TOTAL NET REVENUES”: 8,475 / 6,471 / 14,352 / 22,625 / 30,482. Row “TOTAL GROSS PROFIT”: 3,560 / 4,166 / 8,532 / 13,133 / 17,157. Row “Total Gross Margin”: 42% / 64% / 59% / 58% / 56%. The sheet carries no currency symbol, no “(Eur)” header and no unit label such as “000s”. ⚠ Read from the page image at 400 dpi, and the estimate-column labels are the reason. All five labels carry a space before the letter on the sheet. This sheet’s extraction drops that space in two of the five, and an earlier version of this footnote reproduced the extraction’s spacing in those two — the same correction [^licensing] records for sheet 7, applied here, where sheet 7 served as the control. https://epstein-data.com/EFTA00521863 p. 17. ↩ ↩2 ↩3

  20. Same document, sheet 18, headed “OPERATING EXPENSES AND EBITDA:”. ⚠ Read from the page image at 400 dpi, and the heading is the reason. The heading is complete on a single line at the head of the table, with a terminal colon; it is not a continuation of the preceding sheet’s heading, which is a different heading. It is the extraction that splits it, carrying “OPERATING EXPENSES AND” at the top of the stream and “EBITDA:” twenty-three lines further down, below every row label, because the heading cell and the column block are lifted as separate runs. An earlier version of this footnote took that split for the sheet’s state and described the heading as continuing. Column labels on the sheet: “2009”, “2010 F”, “2011 E”, “2012 E”, “2013 E”. Row “TOTAL COSTS”: (14,606) / (9,489) / (10,037) / (11,090) / (12,140). Row “EBITDA”: (11,046) / (5,323) / (1,505) / 2,043 / 5,017. Row ”% of Net Revenues”: -130% / -82% / -10% / 9% / 16%. Parentheses are the sheet’s convention for negatives. This sheet likewise carries no currency or unit label. https://epstein-data.com/EFTA00521863 p. 18. ↩ ↩2 ↩3

  21. E-mail forward, EFTA02320837, a single sheet stamped EFTA_R1_01230482: produced header “Sent: Sunday, January 23, 2011 5:52 PM”, “To: Jeffrey Epstein”, “Subject: Fwd: Part 2”, “Attachments: Info Memo - 19th jan ‘11 - Part 2.pdf; Untitled attachment 00043.htm”. Beneath: a “From:” line naming Asim Abdullah, a “Date:” line reading “Januar 23 2011 6:44:20 PM GMT+01:00”, and a “Subject:” line reading “Fw: Part 2”, above the adviser’s “Sent: Wed Jan 19 17:03:54 2011” above “Subject: Part 2”. The month name is damaged in this extraction. No sender is supplied here, and no statement is made about the state of those fields: they were not read at the image. Text layer only. https://epstein-data.com/EFTA02320837 p. 1. ↩ ↩2 ↩3 ↩4 ↩5

  22. Same document, sheet 16, headed “Attractive Assets and A Particularly Clean Balance Sheet”: a flagship store “spreads over some 360 sqm (or some 4,000sqft)” with “a significant key money value, estimated at Eur9.5m about 3 years ago by a leading commercial real estate agent in Paris”; “c. 730 sqm of prime offices and showrooms above the flagship store on Avenue Montaigne”; ownership of “the DIVA brand name in the perfume and cosmetics category”; and “a significant tax-loss carry-forward (c. Eur80mn), which can be offset against future profits of the business or similar activities”. The final line is “The company is virtually clear of bank debt.” ⚠ Read from the page image at 400 dpi, and the image corrects the tax-loss figure. This sheet’s extraction substitutes a capital letter for the zero inside that figure; the sheet itself prints a zero, cleanly. An earlier version of this footnote quoted the extraction’s damaged form as the sheet’s own and the body of this page said the sheet rendered it that way, which was a statement about the text layer and not about the paper. The damaged token is not reproduced here. ⚠ One measurement on this sheet is taken and not acted on: at two inter-word positions the gap between words measures 58 px at 400 dpi and 88 px at 600 dpi, where every ordinary word space on the sheet measures 29–46 px at 400 dpi (n=132 over 15 lines); the extraction records a single space at both. The anomaly is on the paper and reproduces across both resolutions, but this page has not established this face’s space advance, so whether the sheet sets one wide space or two is not settled here and the quotations above are left as single-spaced. https://epstein-data.com/EFTA00521863 p. 16. ↩ ↩2 ↩3

  23. E-mail forward, EFTA01809430, a single sheet stamped EFTA_R1_00167842. The produced header is damaged in this extraction. The quoted message beneath reads “Sent : Tue Jan 18 09:07:16 2011” above “Subject : Book completed”, and the sign-off block gives an office, a legal form and a company web address, with labelled contact fields whose values are not reproduced here. The forwarding line reads “Sen ing more e ai s as you requested…”. Text layer only; this page makes no statement about how any field on this sheet was treated by any producing party. https://epstein-data.com/EFTA01809430 p. 1. ↩ ↩2

  24. The producing party’s Sent: header is UTC on this document family. Established on three exact same-message pairs set out in the table in the body of this page, each matching a produced stamp against a copy of the identical message carrying an explicit +0000 offset, to the second: EFTA02002690 against EFTA00679679; EFTA01833847 against EFTA00679596; EFTA01834590 against EFTA00903061. A fourth, weaker check is consistent: a forward produced at 5:49 PM carries a quoted GMT+01:00 stamp of 6:44:05 PM, that is 17:44:05 UTC, four minutes and fifty-five seconds earlier. ⭐ A fifth check is independent of every header on the page and is the strongest of them: EFTA02534817 p. 1 prints the machine field date-received 1295807230, which is 18:27:10 UTC on 23 January 2011 and 10:27:10 at an offset of minus eight hours, matching both the produced stamp and the quoted one to the second. Unzoned quoted headers are not converted anywhere on this page and no interval is computed across one; the one interval on this page that runs between two zoned stamps in different zones, the hour and twenty-nine minutes of 22 January, is set out with its conversion, its offset and its sign control in the body. https://epstein-data.com/EFTA00679679 p. 1. https://epstein-data.com/EFTA02534817 p. 1. ↩ ↩2 ↩3 ↩4

  25. The sentence “Great the owner of ungaro is coming at 830” stands in six distinct documents in this release, one sheet each: EFTA00679596 p. 1, EFTA01798568 p. 1, EFTA01833847 p. 1, EFTA02375908 p. 1, EFTA02535083 p. 1 and EFTA02686633 p. 1. Counted by a column-scoped full-text query over the corpus and reconciled against the 67-document partition set out in the table above. ↩

  26. Information memorandum, EFTA00521863 sheet 3, headed “An Acclaimed New Designer”: “A young, up and coming designer, acclaimed by the international press, has recently taken the creative helm”; “Giles Deacon new designer since summer ‘10”; “Oct ‘10: presentation of first collection (SS ‘11) by Giles Deacon. Euphoric reception by the press.”; “This first collection is clearly starting to re-entice retailers to the brand”. Read from the page image at 400 dpi, where all four quoted strings stand as quoted. A fifth bullet, on the designer’s use of the brand’s archives, stands on the sheet and is not quoted here; each string above is a complete bullet or the complete lead sentence, so nothing is elided inside a quotation. https://epstein-data.com/EFTA00521863 p. 3. ↩

  27. Same document, sheet 4, headed “Enthusiastic Reception By The Press”, reproducing four press extracts beside a portrait captioned “GILES DEACON”. The first is attributed on the sheet to Booth Moore over a Los Angeles Times masthead and includes “Hopefully, he can inject some cool into Ungaro too.” The second is attributed to Suzy Menkes over a blackletter New York Times masthead and reads “By contrast, Mr. Deacon is a 40-year-old designer with a couture sensibility who has developed his own house and could make a serious job of resuscitating the Emanuel Ungaro heritage…”. A third carries a VOGUE.COM mark and is attributed to Giles Deacon himself; a fourth carries a Harper’s Bazaar mark. Read from the page image at 400 dpi, and the image is the point of this footnote. The extraction of this sheet renders the Menkes sentence’s final verb as a broken token, and renders the two mastheads as “CosAngeles times” and “fork attics”. ⚠ An earlier version of this page printed the extraction’s broken token inside quotation marks as the critic’s own words and stated that the sheet was damaged. The sheet is not damaged: at 400 dpi the word is “resuscitating”, complete and clean, and the count of extracts was given as three. Both were corrected against the image. https://epstein-data.com/EFTA00521863 p. 4. ↩

  28. The designer enquiry stands in 26 distinct documents in this release. Twenty-four are reached by a corpus phrase search for the question itself; two further copies are reached only by the house’s name, because quoted-printable encoding damages a word inside the sentence in those two extractions. The damaged token is not reproduced here. ⚠ They are not one sheet each and they are not copies of one message. The 26 documents stand on 49 sheets in total, of which 26 carry the house’s name and so enter the partition above; and of the twenty-four reached by the question’s own words, twenty-one carry more than one message stamp and are chains in which the question is quoted, while three are the single message alone. One of those three carries none of the three boundary markers and so cannot be placed by the marker count at all: it is a single sheet of 284 characters on eleven lines, carrying, in this extraction, two header-shaped labels — a From label and a Subject label carrying a quotation marker, both of them stripped of their values — and one occurrence of the question. This sheet was not rendered, so nothing is asserted here about where either label stands on the paper. It is counted here as a single message on the absence of any boundary marker, not on the label count and not on a stamp, and that classification is not settled by the labels either way.

    The arm, as it was actually run. Counted in DOCUMENTS by column-scoped full-text queries over the corpus, in SHEETS by a row count over the same documents, and in MESSAGES by counting boundary markers over the whole document, not over the sheet carrying the question, without reading any body. The marker arm is Date or Sent at the start of a line with or without a colon, plus On … wrote. ⚠ The colon is not required, and that is load-bearing rather than incidental. In this population the extraction repeatedly splits a produced header’s label from its value onto separate lines, leaving Sent alone on a line above a date. Eight such headers stand across these documents, and they are genuine message boundaries that a colon-bearing arm does not reach. This page’s own footnotes quote that colon-less produced form four times. ⛔ What the arm also matches, and should not: a line-initial device-signature footer of the form Sent using …, which is not a message boundary and which occurs 42 times across these documents; the chain count is unaffected because every document carrying one carries a real boundary too.

    The grid, so the figure can be tested. Chains, meaning documents carrying two or more boundary markers, out of the 24 reached by the question’s own words:

    line-local On … wroteOn … wrote may span a newline
    colon required, whole document1920
    colon optional, whole document2121
    colon optional, question’s sheet only1919

    The figure of 21 is the colon-optional, document-scoped count and is stable under both readings of the On … wrote marker; it is also unchanged by whether Date/Sent is anchored to the start of a line. A colon-bearing arm returns 19 or 20 and a sheet-scoped arm returns 19, so neither reproduces it. ⚠ An earlier version of this footnote printed the marker as "Sent:" and did not state the scope, so the label described a query that does not produce the number beside it; the number was arrived at by reading the documents’ structure and is unchanged. Reconciled against the 67-document partition. https://epstein-data.com/EFTA02014303 p. 1. ↩

  29. The two zoned stamps inside the forwarding chain are “January 23, 2011 6:44:05 PM GMT+01:00” at EFTA02375967 and “6:44:20 PM GMT+01:00” at EFTA02320837, fifteen seconds apart, being the owner’s two forwards of parts 1 and 2. Converted: 17:44:05 and 17:44:20 UTC. Both quoted headers carrying an explicit offset. Text layer only. https://epstein-data.com/EFTA02375967 p. 1. ↩ ↩2 ↩3 ↩4

  30. E-mail forward, EFTA02034543, a single sheet stamped EFTA_R1_00546293, 352 characters in the extraction. ⚠ Read from the page image at 600 dpi, and the image changes two things this page previously took from the extraction. First, the produced header is not damaged on the sheet: it reads “Sent: Sun 1/23/2011 6:03:58 PM” in full, above “Subject: Fwd: Information Memorandum - part 1”. It is the extraction that breaks the date, rendering it “Sent un 11 6:03:58 PM”, and an earlier version of this footnote treated that breakage as the document’s state and declined to take a date from it. The stamp is a produced Sent: header on this family and so is 18:03:58 UTC on Sunday 23 January 2011. Second, the quoted body reads “From: Asim Abdullah” above “Also ▮.I am in paris till fri but can travel to ny on my way to sf if Jeffrey wants me to meet someone there…”, where ▮ is a producer’s mark on the sheet; the extraction renders that region ”.M.”, and the earlier version of this page printed an ellipsis there, which turned a withholding into the writer’s own punctuation. Nothing about what the mark covers is stated or inferred here, and no other copy is consulted or compared. https://epstein-data.com/EFTA02034543 p. 1. ↩ ↩2

  31. E-mail forward, EFTA02687313, a single sheet stamped EFTA_R1_02014618: “From: Asim Abdullah”, “Sent: Sunday, January 23, 2011 6:30 PM”, “Subject: Fw: Book completed”, “Attachments: FILE PERSONNEL COSTS AVEC ANCIENNETE 2010-2012.xls; CASH FLOW 3 YEARS PLAN.xls; BOOK 3 Years’s plan.xlsx”, body “Sending more details as you requested…”. The quoted message beneath is the 18 January internal note at “Sent: Tue Jan 18 09:07:16 2011”. None of the three workbooks is in this release; only their names, in this header. Text layer only. https://epstein-data.com/EFTA02687313 p. 1. ↩ ↩2

  32. A copy census of Epstein’s 23 January message, run as a column-scoped corpus query on the opening words and reconciled against the partition above. The release holds six copies: four in which it is quoted beneath the owner’s reply, at EFTA02002690 p. 1, EFTA01791414 p. 1, EFTA01834620 p. 1 and EFTA00679679 p. 1, and two as sent, at EFTA01834414 p. 1 and EFTA02534817 p. 1. ⭐ The census, run over all six text layers with whitespace normalised, and it is stronger than this page first claimed. Five of the six print “n;t”: every copy except EFTA02002690, whose extraction gives a different character in that position. Five of the six print “licencing” with a c. The middle token is the unstable one: three extractions give “beed” and the other three each substitute a different character inside it, which is ordinary character-substitution damage and the substituted forms are not reproduced here. Two copies, EFTA01791414 and EFTA00679679, carry the whole transcribed passage verbatim in the text layer alone, and either may be cited as a corroborating pin for it. An earlier version of this footnote said that no two extractions agree, and that was wrong. The transcription in the text of this page was read from the page image of EFTA02002690 at 600 dpi and is confirmed at the image of EFTA02534817 at 400 dpi; a checker binding it against EFTA02002690’s text layer will still report it unresolved, because that is the one copy whose extraction differs, and no word of the quotation should be altered to satisfy it. ↩

  33. Epstein’s message as sent, EFTA01834414 p. 1, a single sheet stamped EFTA_R1_00213472, 1,088 characters. It is not a reply and quotes nothing. The producing party’s header reads “From:” above a damaged rendering of the sender’s name and “Sent Sun 1/23/2011 6:27:10 PM”; the recipient field is not reproduced here. The body is the same message quoted in the text of this page, rendered in this copy with a different character substitution in the word before “working” and in the surname in the fourth clause. This is the copy that fixes the message at 18:27:10 UTC and so resolves the quoted 10:27:10 header as US Pacific Standard Time, an exact eight-hour offset with the seconds agreeing. Text layer only. ↩ ↩2 ↩3

  34. A second copy of the same outgoing message, EFTA02534817 p. 1, a single sheet stamped EFTA_R1_01679929, 1,227 characters: “From: Jeffrey Epstein” with the sender’s address, which is a contact-class value and is not reproduced here, and “Sent: Sunday, January 23, 2011 6:27 PM”. It corroborates the 6:27 PM stamp to the minute, independently of the preceding copy. ⭐ It also carries, in a trailing line of mail-client metadata, the machine field date-received 1295807230, a Unix epoch value equal to 18:27:10 UTC on 23 January 2011, which is 10:27:10 at an offset of minus eight hours. That line is printed on the sheet and was read there, not inferred from the extraction. ⚠ Quoted-printable damage runs through this copy’s body and footer, and it is on the paper, not in the extraction: at 400 dpi an equals sign stands in place of a letter at several points in the sentence and in the confidentiality block. The damaged tokens are not reproduced. This copy’s body also prints “n;t”, “beed” and “licencing” at the image, so the central transcription is confirmed at a second sheet as well as at the five text layers. The recipient field is not reproduced here and no statement is made about its state. Read from the page image at 400 dpi. ↩ ↩2

  35. The same message pair in a copy carrying an explicit offset, EFTA00679679, a single sheet: “From: “Asim Abdullah"", “Subject: Re:”, “Date: Sun, 23 Jan 2011 18:38:50 +0000”, with the identical four notes and the identical quoted reply. Cited here only to establish the zone of the produced stamp. Recipient fields are not reproduced. Text layer only. https://epstein-data.com/EFTA00679679 p. 1. ↩ ↩2 ↩3

  36. The two intervals are computed between a produced header and a produced header, both UTC on this family per the preceding section, so each is zoned against zoned. Part 1 is sent on to Epstein at 17:49 (EFTA02375967 p. 1) and his critique is sent at 18:27:10 (EFTA01834414 p. 1): 18:27:10 minus 17:49:00 is +00:38:10. His critique at 18:27:10 and the owner’s four notes at 18:38:50 (EFTA02002690 p. 1): 18:38:50 minus 18:27:10 is +00:11:40. The sign control on both is that the later act must follow the earlier, and both differences are positive. Neither interval crosses midnight and neither crosses a daylight-saving boundary, US daylight time in 2011 having run from 13 March to 6 November and European summer time from 27 March to 30 October. The date-rollover question is answered by deriving the window from the offset rather than by assuming one: a UTC instant converts back into the previous day only when it falls between 00:00 and one second before the offset’s own hour, so at UTC−5 the window runs to 04:59:59 and at UTC−8 to 07:59:59. Every stamp used anywhere on this page, zoned or unzoned, stands at 09:07:16 UTC or later, so none of them falls in either window and no date on this page turns on a rollover. The one rollover that does occur runs the other way, eastward into Central European Time, and is set out in The clocks. ↩

  37. The same message under a producing party’s header, EFTA01833847, a single sheet stamped EFTA_R1_00212481: “From: Jeffrey epstein”, “Sent Sat 1/22/2011 3:32:32 PM”, “Subject Re: No subject”, body “Great the owner of ungaro is coming at 830”. Cited here for the produced stamp, which matches the zoned copy to the second. Text layer only. https://epstein-data.com/EFTA01833847 p. 1. ↩

  38. The same message in a copy carrying the opposite zone, EFTA00903061, a single sheet: “From: Jeffrey Epstein”, the sender’s address not reproduced here, and “Date: Sat, 22 Jan 2011 23:04:10 +0000”, same one-line body. This copy and the copy at [^deacon] are the same message at the same instant under two dates. Text layer only. https://epstein-data.com/EFTA00903061 p. 1. ↩

  39. The same message under a producing party’s header, EFTA01834590, a single sheet: “From: Jeffrey Epstein”, “Sent Sat 1/22/2011 11:04:10 PM”, same one-line body. Cited here for the produced stamp, which matches the UTC copy to the second. Text layer only. https://epstein-data.com/EFTA01834590 p. 1. ↩

  40. Same document, sheet 23, headed “IMPORTANT LEGAL INFORMATION (5)” above the subheading “Risks”, a six-bullet list. The bullet quoted is headed “Potential losses.” and reads in full “Emanuel Ungaro has been reporting losses in recent years. Whilst results have been improving and management is looking for the Company to reduce losses, and then to be profitable in future years, it is worth mentioning that, should these losses continue, they would have to be financed.” The other five bullets are headed “Difficult economic environment.”, “Trading down.”, “Fashion and Beauty Trends.”, “Higher fixed costs and working capital requirements associated with reintegrating licenses internally.” and “Credit conditions.” Read from the page image at 400 dpi: the list is six bullets, the count is as stated, and the quoted bullet and all six headings stand as quoted. https://epstein-data.com/EFTA00521863 p. 23. ↩

  41. Same document, sheet 14, headed “Restructuring Program Under Way”, six bullets: “Key management positions (Chief Executive Officer and Chief Financial Officer) are available to be filled by the future shareholder/operator.”; a bullet on the Chief Operating Officer “promoted from within the company some 1and1/2 years ago” with “over 20 years within the company”; a bullet on a cost overhaul “in the last 12 months”; “Only c.25 staff members”; “Rich archives”; “Strong and well-protected IP portfolio”. Read from the page image at 400 dpi: the list is six bullets, the count is as stated, and every quoted string stands as quoted. https://epstein-data.com/EFTA00521863 p. 14. ↩

  42. Same document, sheet 15, headed “2010 Organisation Chart”, showing “Asim Abdullah” above “Président & CEO” — the sheet prints the French form with its accent, which this extraction drops — and that above a General Manager, with boxes for the women’s and men’s ready-to-wear studios, the atelier, sales, communications, the Montaigne flagship, an advertising and events function, finance, accounting, human resources and general services. ⚠ Counted at the image: sixteen boxes, of which thirteen carry a headcount and three carry none — the two boxes at the head of the chart, which carry names and offices instead, and the finance department box, whose three sub-boxes carry the headcounts. The thirteen stated headcounts total 39. An earlier version of this footnote said each box carried a headcount. The sheet carries the word “CONFIDENTIAL”. The General Manager is named on the sheet; this footnote refers to that person by office, while the name stands elsewhere on this page inside quotations of two other sheets, as Who is named here, and who is not records. This is one of the two sheets on which the extraction damages the wordmark, rendering it “manual ungao”, so the sheet is not reachable by a search for the house’s name. Read from the page image at 400 dpi. https://epstein-data.com/EFTA00521863 p. 15. ↩

  43. FashionUnited, “Ferragamo sells Ungaro brand”, dated 2005 on the page, at https://fashionunited.uk/news/fashion/ferragamo-sells-ungaro-brand/2005090436144 , fetched 18 October 2026, HTTP 200, 64,776 bytes, content-type text/html. The article states: “Emanuel Ungaro said Thursday that Gruppo Ferragamo signed an agreement to sell 100 percent of the company to San Francisco-based high-tech entre pr eneur Asim Abdullah. Terms of the deal were not disclosed.” It further states that the Ferragamos “will continue to hold the Ungaro licenses for leather goods and fragrances”, and that “A WWD source estimated the deal would be worth $5 million to $10 million for the fashion and fragrance businesses.” The word breaks inside “entre pr eneur” are in the served page. One divergence from the memorandum is noted rather than resolved: this article says “Salvatore Ferragamo bought Ungaro in 1996”, where the memorandum’s ownership slide gives the Ferragamo period as “‘95-‘05”. Each is attributed to its own source and neither is corrected by the other. ↩

  44. A contemporaneous trade report of the September 2012 Aeffe licence was sought at https://ae.fashionnetwork.com/news/Italy-s-Aeffe-adds-Emanuel-Ungaro-to-brand-portfolio,282524.html on 18 October 2026. The fetch returned HTTP 403 with 5,650 bytes of an interstitial page reading “Just a moment… Enable JavaScript and cookies to continue” and no article text. This is recorded as a failed fetch and not as an absence: nothing about the 2012 agreement is asserted on this page from that source, and the only statement made about it here is the one carried by the 2015 report that was successfully fetched and read. ↩

  45. A count over the 34 sheets of the two runs. 24 sheets contain an extractable occurrence of the token “ungaro”; 10 do not, because the house’s wordmark is set in a rounded lowercase display face that the extraction resolves differently almost every time. Forms it produces across the run include “emanuel ungera” (EFTA00521855 sheet 3), “emanuel ung€ro” (EFTA00521855 sheet 5), “emaruel ungaro” (EFTA00521863 sheet 6), “emanual Lngaro” (sheet 7), “Acreit.ial ungaro” (sheet 9), “ernanuel ungei•o” (sheet 10), “ernenual ungaro” (sheet 18) and “ernarual ungaro” (sheet 25); each was located by querying the two runs sheet by sheet, and none of the eight stands on the cover sheet this footnote previously pinned them all to. Two further forms, “ennanual ungara” on sheet 13 and “manual ungao” on sheet 15 of EFTA00521863, are the only hits returned by two single-character variant arms run over the whole corpus, and neither reaches any document outside this run; an earlier version of this footnote called them “two of these forms” while listing neither. None of these forms is a name of the house, none is reproduced as one, and none appears in this page’s aliases: field. Counted at the text layer over all 34 sheets; the wordmark itself read at the page image at 400 dpi on the cover and on sheets 1, 2, 4 and 7. ⚠ A pin checker will report several wrong rows against this footnote and they are an artefact of its structure, not a mispin: the footnote quotes ten damaged forms that stand on eight different sheets while carrying three page pins, and the checker pairs every pin on the line with every quotation on the line. Each form’s own sheet is named beside it above. No pin here should be moved on the strength of those rows. https://epstein-data.com/EFTA00521855 p. 1. https://epstein-data.com/EFTA00521863 p. 13. https://epstein-data.com/EFTA00521863 p. 15. ↩

Backlinks