Greek Yields: "Certain Default, But Not Yet"

John P. Hussman, Ph.D.
All rights reserved and actively enforced.
Reprint Policy

Over the past few weeks, we've observed a fairly abrupt shift toward weakening across a variety of economic indicators, coupled with a deterioration in market internals. Investors have also placed a great deal of attention on the prospect for a default in Greek government debt. From our perspective, it is important to separate these issues. At least over the near term, the likelihood of a Greek default is quite small. The more pressing issue for investors is that market internals have deteriorated measurably, signaling a shift toward risk aversion among investors in an economy where the scope for further fiscal and monetary intervention is limited, and in a market that remains overvalued on the basis of measures that are actually well-correlated with subsequent market performance.

As usual, I have to emphasize the importance of using historically reliable valuation measures, as opposed to the common but wholly unreliable "forward operating earnings times arbitrary P/E" approach that is the mainstay of market commentators who seem to view historical research as optional. Based on a variety of historically reliable measures (detailed in numerous prior comments), we currently estimate that the S&P 500 is priced to achieve 10-year total returns averaging only about 4.1% annually, which (as a result of the recent market decline) is somewhat higher than the 3.4% we estimated several weeks ago, but remains unsatisfactory from the standpoint of investment merit.

I am not suggesting that the prospect of Greek default is unimportant, but based on the profile of Greek yields across the maturity curve, concerns about a near-term risk of Greek default appear premature. My impression is that a Greek debt default (and an exit from the European Monetary Union) is much more likely to occur in the final phase of the next bear market - not as the opening salvo. The bulk of investor attention at present should be placed on valuations, market internals, and economic factors.

The market is clearly oversold on a short-term basis (though only slightly on an intermediate-term basis). This has reached the point where some good news about Greece could prompt a relief rally, but the quality of market action in any advance will be important. We could see a bit of latitude to accept modest exposure on the basis of speculative merit, but that would require a shift toward more favorable internals, and it's likely that any material advance would quickly reestablish an overvalued, overbought, overbullish, rising-yields syndrome. An improvement in market internals from lower levels would provide greater latitude for sustained exposure.

Despite the short-term oversold condition of the market, I should be clear that we are presently observing a combination of evidence that is typical of early bear markets - having some potential to be reversed, but with a generally dangerous record overall. This evidence includes the present combination of unfavorable valuations and unfavorable market action, developing concern from the most accurate version of our recession warning composite (which would be completed with a monthly S&P 500 close below about 1250 and another weak ISM report), a recent advance that has already passed the historical norms for extent and duration of cyclical bulls within secular bears (see Hanging Around, Hoping to Get Lucky ), and the neutral intermediate-term but hostile longer-term evidence we observed at the early May peak (see Extreme Conditions and Typical Outcomes ). All of this presently holds us to a generally defensive investment stance.

"Certain default, but not yet"

The following chart presents the probability of a Greek default implied by the yields on Greek debt of varying maturities. Each line assumes a different "recovery rate," which is the proportion of face value that investors would actually recover in the event of default. The way to think about this is that a given default premium could either compensate for a high probability of a small default, or a small probability of a major default. We can't observe the market's recovery assumptions directly, but we can easily estimate the range of possible combinations of default risk and recovery implied by Greek yields.

A few things are important to note. First, even with yields on short-term Greek bills pushing toward 20% annualized, the implied probability of a near-term default is very low. Indeed, the only way that Greek yields would be consistent with even a 50% chance of default in the next 3 months is to assume a 95% recovery rate. That sort of default is highly unlikely, because it would do nothing to materially alter the overall debt burden. For plausible recovery rates, the implied probability of a near-term Greek default is less than 7%. Simply put, despite the strained political and social situation in Greece, the bond market is demonstrating very little concern about a near-term default. At the point that a near-term default becomes likely, we would expect to see one-year yields spiking toward 40% and 3-month yields pushing past 100% at an annual rate (essentially pricing near-term bills toward the anticipated recovery rate).

This temporarily reassuring situation, unfortunately, strongly contrasts with the longer-term outlook for Greek debt. Even assuming a 60% recovery rate (that is, assuming a default would wipe out 40% of the Greek debt burden), the implied probability of a Greek default within the next two years is effectively 100%. The only way you get a lower probability is to assume a far lower recovery rate.

Indeed, just to get the 20-year default probability below 100%, one needs to assume a recovery rate of less than 20% of principal. And even if you assume a nearly-complete wipeout with only 5% recovery, you're still left with a 20-year default probability of greater than 85%. So yes, we can probably assign less than 100% probability to defaults that would have disastrously low recovery rates, but if we stay in a plausible range of recovery rates, Greek yields suggest that we will approach a near-certainty of Greek default by mid-2013. From a time perspective, the only way Greek yields are consistent with less-than-certain default by the end of 2013 is if we assume recovery rates of less than 60%.

In short, either long- and intermediate-term Greek debt is a tremendous bargain here, or it is going to default. Unfortunately, the fiscal situation would almost inescapably require other European countries subsidize Greece for decades to come in order to avoid a debt restructuring. Taken together, the evidence surrounding Greece screams "Certain default, but not yet."

[Geek's note.. and in this case, Greek's note too: The implied probability of default is calculated as (1-exp(-(Y-Yf)*T))/(1-R), where Y is the yield in decimal form, Yf is the yield on a presumably default-free security (in this case, we're using comparable German yields), T is maturity, and R is the recovery rate, also in decimal form.]

Market Climate

As noted above, an ensemble of other factors may allow for modest, transitory market exposures in proportion to the return/risk profile we estimate, but the estimated return/risk profile of the market is not likely to improve markedly until we see either a significant improvement in valuations, or a firming of market interna ls that does not also re-establish an overvalued, overbought, overbullish, rising yields syndrome. For now, we remain defensive, with only minor latitude for deviations from a tight hedge.

In bonds, the Market Climate last week was characterized by unfavorable yield levels and mixed yield pressures. Though a significant credit event or other flight-to-quality could push yields somewhat lower, investors in long-term Treasury securities have to weigh the fact that an increase in yields of 20-40 basis points would wipe out the yield advantage of these securities over T-bills. This is a particularly challenging problem for the Federal Reserve, which has now leveraged its balance sheet 53.5-to-1. It's also worth noting that the Fed's SOMA portfolio passed $2.6 trillion last week. As Fed noted in its report Domestic Open Market Operations during 2010 :

"With progress towards its statutory objectives of maximum employment and price stability disappointingly slow in the fall of 2010, most Committee members judged it appropriate to provide additional monetary accommodation. Accordingly, the FOMC announced at its November meeting that it intended to increase the total face value of domestic securities in the SOMA portfolio to approximately $2.6 trillion by the end of June 2011 by purchasing a further $600 billion of longer term Treasury securities in addition to any amounts associated with the reinvestment of principal payments on agency debt and MBS."

So aside from a few table scraps, QE2 is effectively complete, and the continuing issuance of Treasury debt to finance the ongoing federal deficit - about $1.5 trillion for 2011 - will now have to be absorbed by the U.S. public and international lenders. It's easy to see how $600 billion of Fed purchases over an 8-month period has made that math a bit easier for the private markets. Frankly, it's difficult to envision all of this playing out smoothly from the extremes we've now established.

In Strategic Total Return, we reduced our duration in Treasuries last week to just 1.5 years, and slightly increased our exposure in precious metals shares (still about 18% of net assets), where we estimate an unusually strong return/risk profile. Precious metals shares do carry significant volatility, however, so the "risk" aspect should be recognized. As I noted last week, further weakness in the ISM measures, continued inflation pressure, and various other factors could push our exposure somewhat higher, but we've rarely our sustained precious metals allocations beyond 20% for long. I view our present exposure to precious metals shares as significant, and it is in proportion to the return/risk profile that we presently estimate.