27 March 2015 US Fixed Income Weekly explains for about 35 bps of the excessive flatness using the Michigan beta in our model. If we plug in the lower of these two market-based measures, without recalibrating the model, 5s-10s looks 20 bps below fair value. Another explanation could be that there is some upward pressure on the 5y point of the curve. This could be the market pricing in a Fed that seems to be drifting apart rather than coming to consensus on the appropriate terminal Fed funds rate. In the March FOMC projections, the most hawkish and the most dovish members differed on their projection of the long run policy rate by 125 bps. Back in December their difference of opinion was 100 bps. One year ago, that difference was just 75 bps. Difference between highest and lowest projections of 'Fed ACM term premium and 2y-5y- I Oy term premium fly longer run policy rate 1.75 Percent 130 1 1.25 1.00 1 0.75 -I Ono o.n 0.00 Holm t st ii-Ocit tente gatn cloy 1008 2009 2010 2011 2012 2013 2014 2015 —2y term premium (%lett stele) —Sy term premium (%, left scale) FOMC Protections Release Date — soy term Prenliati it left scale) —1Y-5y-10Y Ay LW rithl WIN Scarce federal Raw/were twin., lint Seine federal Amer., *no 'Dania ant The second chart on top shows the Fed's ACM Treasury term premium for the post-crisis period. Note that even though term premium on all parts of the curve have declined, only in recent months the 5-year term premium had begun trading on top of 10-year term premium. By the Fed's definition, this means investors are now more anxious about - and thus demanding more compensation for - the risk that short rates do not evolve as expected in 5 years time than in 10 year's time. The 2y-5y-10y term premium fly also illustrates the point that required compensation for term risk is now the highest in the 5y sector relative to other parts of the curve. The fly residual suggests that this factor accounts for another