During February of 2011, PIMCO announced that it has sold its portfolio of U.S. Treasury bonds in anticipation of higher yields due to inflationary concerns arising from a second round of quantitative easing by the Fed, known as QE2. Expectations for lower bond prices (TLT) not only did not materialize, but instead yields fell straight down rapidly, as the monthly chart of the 10-year bond yield below shows:
By the end of 2011, the 10-year note yield was at 1.87% and the 30-year bond yield was at 3.00%, having been at 3.31% and 4.26%, respectively, at the start of the year. In hindsight, the PIMCO decision was bad and had an impact on their performance. The PIMCO Total Return Fund (PTTRX) gained 4.26% for 2011, a performance much lower than the Barclays U.S. Aggregate Index, which returned 7.84%. Some would think that this was a bad performance given that an investment in TLT would have generated in 2011 a total return close to 30% (including dividends), while stocks (SPY) finished almost flat, as the relative performance of SPY and TLT for 2011 shows:
In Defense of PIMCO
However, I will here defend PIMCO"s strategy for 2011, because an analysis based on hindsight is not appropriate. Fortunately, or unfortunately for that purpose, and depending on how one conceives that, an after-the-fact analysis makes things clear. But before the fact, a lot of things are usually at stake. What about if QE2 had really caused yields to start to rise? Due to their size alone, PIMCO would have caused yields to rise even further should they have decided to sell late, even causing an interest rate rise with potential impact on economic growth. But more importantly, the primary objective of a fixed income manager is to preserve capital. This is what some equity, commodity and forex speculators do not understand. A +4.26% return is worse than a +10% return but still positive and better than a -10% loss. We are talking about huge numbers in absolute terms here, and a large loss would have impacted a great number of conservative fixed income investors, like a great number of pension funds. Thus, although the decision to sell Treasuries in 2011 appears bad based on hindsight, in my opinion, it was an optimal decision from a game theoretical view point. For example, yields are not affected only by real macroeconomic factors but also from investor psychology. What about if the Chinese had decided to sell a good part of their Treasury holdings first? Such scenario would have caused PIMCO to chase the market and sell at much lower prices resulting in their first ever loss. In decision theory, it is sometimes better to take a small loss temporarily in order to secure gains in the future. This is what PIMCO did in my opinion and in this respect its strategy was optimal. Optimality greatly depends on the objective considered.
PIMCO's New Bet
According to an article in Forbes, by the end of August 2012, PIMCO has cut its holdings of Treasuries to 21% from 33% and has increased its holdings of municipal bonds. At this point, I must admit I know little about municipal bonds other than what Meredith Whitney has forecasted about them. When I worked with bonds, it was mainly with Treasuries and corporates. It is clear though that lately municipal bonds have attracted a lot of interest. John Coumarianos of Institutional Imperative wonders whether Meredith Whitney was early in her forecast, but it appears that given the vigilance of Fed, this may not be important any longer. Will QE4 also include Muni bonds?
Historically, municipal bonds have yielded 10%-15% less than Treasuries. However, after the financial crisis, this relationship reversed significantly for the first time due to investor concerns about increased risk and flocking to Treasuries, as John Waggoner writes in USA today. That in turn has pushed Treasury yields well below municipal bond yields, as the charts below show:
It may be seen from the above charts that the new bet of PIMCO is similar to a mean-reversion trade. Actually, they are betting that at some point the yield spread between munis and Treasuries will close and approach average levels of the past. Obviously, there is no trade without risk.
What could go wrong with this bet
- The spread may close but at higher yield levels for both treasuries and municipal bonds.
- More muni defaults make the spread to open further while the economy does not pick up, keeping Treasury yields low.
A question with the above two scenarios is where the stop-loss of this bet is placed, something that only PIMCO knows.
What factors will influence the profitability of the bet?
- The economy improves, Treasury bond yields rise and municipal bond yields fall due to increased investor interest.
- The economy gets a lot worse, QE4 is instituted, and it includes municipal bonds.
Thus, it may be seen that, according to the above analysis, the bet on municipal bonds may work well with the extreme scenarios of either a much improved economy or a much worsening economy with a need of a broader bond purchasing program by the Fed. If the economic situation remains unchanged with even some scattered reports of muni defaults, this bet may not work out very well, again according to this analysis and with the limited number of parameters considered.