Many investors were surprised by the hawkish comments in the December Minutes of the FOMC.
"A few members expressed the view that ongoing asset purchases would likely be warranted until about the end of 2013, while a few others emphasized the need for considerable policy accommodation but did not state a specific time frame or total for purchases. Several others thought that it would probably be appropriate to slow or to stop purchases well before the end of 2013…"
I took this opportunity to update my US 10-year treasury yield model, based on traditional factors:
1. The level of current Fed Funds rates;
2. The 1-year return of the SP 500;
3. The ISM (positive impact of better economic prospects on yield); and
4. The spread between 5-year and 1-year inflation expectations (a proxy of the change in the inflation premium embedded in the long end of the curve).
I did not add a "dummy" data series (ad hoc series) for episodes of Quantitative Easing. Interestingly enough, a dummy would have been required, not for any of the QE or Twist episodes, but for the debt ceiling negotiations during summer 2011 when US Treasury yields collapsed to their current range (1.6-2.2%).
In addition, the model is estimated with monthly data since 2004 and is based on traditional macro drivers, not on the nature of monetary policy. The chart below (left) can therefore be considered a QE-free model of US 10-year interest rates.
The result is compelling: the UST 10-year yields are at a level slightly above the lower bound of fair value. Thus, the risk of a sharp selloff should be limited. The chart (right) uses the same factor to assess the 3-month change in 10-year UST yield. It shows that the recent increase is also consistent with traditional drivers (slightly above the upper bound, which would suggest a forthcoming decline, or at least that the market has overreacted to the Fed's Minutes…).
Where could we go from here?
Assuming a 10% annual return for the SP 500 in 2013, a slight uptick in 5-year inflation expectations, Fed Funds remaining at 0.25% throughout the year, and an average ISM of 53 in 2013, the model suggests that UST yields fair value would stand at 2.9%.
You may have read this before. For years the so-called bond-vigilantes were supposed to bring rates much higher. This time things are different though:
1. The Fed's new mandate (conditional inflation targeting) and the internal dissensions on the conduct of quantitative easing are clearly less dovish;
2. Twist was the only time when the Fed succeeded in lifting stock prices without rising long-term rates. Any new purchase programs would be more like QE1 and QE2 with the associated risks on the USD and UST yields;
3. The fiscal cliff was avoided… but the deficit will be higher (6% instead of previously forecasted at 5.4%). More bond supply in a context of uncertainty on monetary policy could play a role (the impact will be limited, as QE will run at least until mid-2013). Above all, growth prospects did brighten with the ATRA, which is also negative for yields;
4. From a European perspective, the reduced risk of a euro breakup would make Bund yields less attractive.
It is true that the huge amounts of liquidity provided by central banks reduced the yield arbitrage/competition, as investors have been chasing yields everywhere (see below the sharp fall in credit market yields).
Everywhere, that is, except in the equity market. As can be seen on the chart below, there is still major upside for stocks before they can be deemed "rich" (on a short run basis).
It is clearly too early to call for a return of the "bond vigilantes." Yet, better growth prospects combined with growing uncertainties on the conduct of monetary policy and the expansion of Central Banks' balance sheet, suggests that the potential for higher 10-year US rates is significant in 2013.