Signs of a Double-Dip Emerge

Updated on

 

This was sent by one of our friends:

The global recovery has taken some hard stumbles over the past few months, but hasn’t fallen. Sovereign debt crises in the US and Europe, as well as economic softness in both, contributed to significant market declines. Debt worries on the European periphery have turned into fears of contagion to core countries such as Italy and France. And while the US debt ceiling show-down has been postponed, GDP of 0.4% in Q1 and 1.0% in Q2 raise concerns that the US is falling into a second recession.
Handicapping the risk of a double-dip is difficult. Some economists posit a “stall theory,” in which GDP below 1% foreshadows a recession. This criteria was fulfilled in 82% of post-war recessions, but also gave false positives on at least 7 non-recessions. Recent weak GDP therefore points only weakly to a new recession, particularly since the impairment due to Japan’s Dai-ichi disaster is likely to reverse itself in H2. The more disconcerting possibility is that uncertainty and flagging confidence precipitate a reinforcing downward spiral. Consumer sentiment is an unreliable indicator of recessions (eg, the steep decline in 2005 was misleading), but the combination with heightened uncertainty presents a real risk. Many economists peg the risk of a recession in the 35-45% range, and we agree that the risks have risen.
Despite the greater macro risks, we feel the investment environment remains attractive for two reasons. The first is that many structured credit markets are already pricing in draconian scenarios. Distressed sectors of the non-agency MBS market, for example, are at prices comparable to Q1 2010, despite sharply lower delinquency transition rates, a more stable housing market, and improved bank and consumer leverage. The second is that much of the surprise and confusion surrounding our flagging recovery may be based on incorrect context.
The proper context, in our view, is a recovery from a major banking crisis. Ordinary recessions tend to be followed by strong growth, and the deeper the recession, the stronger the “V” recovery. This observation underpins much of the puzzlement over the present weak recovery. Banking crises, as evidenced by the work of Carmen Reinhart and Kenneth Rogoff, tend to have very weak, prolonged recoveries associated with rising sovereign debt and persistent high unemployment. In this context, recent events are entirely in line with prior experience.
As the Tables below highlight (averages are based on 19 post-WWII and 2 pre-WWI crises), the recent financial crisis looks like a banking crisis not only in the plunge down, but also in the recovery to date. In this context, ballooning debt, sovereign downgrades, and weak growth are normal – and are not by themselves signals of an imminent double-dip. They do highlight real risks, and these risks will likely persist for years. But slipping into a mild recession is a very different proposition than a second banking crisis.
Events that once seemed almost inconceivable in the context of a post-recessionary recovery (such as the US downgrade) are unsurprising in the context of a banking aftermath. Compared to the average banking crisis in Table 2, the US recovery exhibits somewhat higher federal debt growth (90% is in line with historical averages, but US debt will probably climb significantly higher), similar sovereign ratings downgrade, and above-average GDP (buoyed by stimulus). The historical average is a 4.4 year recovery at GDP of 2.1%, compared to 2 years so far in the US, with average GDP of 2.5%. Of course, the US path will surely deviate from the average (in historical crises, recovery times extended to as much as 8 years), but the average provides a baseline from which to consider the US post-crisis recovery. While the decline in US GDP was relatively mild (6.6% versus 9.3% average), similarities in other metrics suggest it will be years before the US recovers broadly.

Rebuilding confidence, restoring credit flow, remediating structural imbalances, and deleveraging are critical to recovery from banking crises, but take time. Confidence and credit availability remain subdued amidst economic and regulatory uncertainty, but structural rebalancing has progressed notably, especially in the housing sector. Homeownership rates have fallen from approximately 69% at the peak to 66%, only 2 points above the 64% stable equilibrium that prevailed from 1985-95. Home prices are roughly in line with long-term price/rent and price/income ratios, and are within 10% of 120-year average real levels.

Deleveraging will likely take years before returning to long-run averages, but significant progress has already been made. Figure 1 shows that the steepest run-ups in debt occurred in the financial and household sectors, but both have declined meaningfully in the past two years. Household debt fell from 99.5% of GDP in Q1 2009 to 89% at present, while financial institution debt fell from 124% to 95%. Much of the decline in financial and household debt was compensated by a rise in government debt1, but even  so total debt (financial, household, government, etc) has declined from approximately 365% of GDP at the peak, to 335% at present. In terms of leverage, consumers and banks are in far better shape than just prior to the crisis (reducing the risk of a severe contraction), but we have a long way to go before pre-bubble levels (reducing the likelihood of a robust recovery).

 Figure 1: The US is Deleveraging in Aggregate

A long, weak recovery is vulnerable to continued threats, and we should expect macro sentiment to continue driving markets. How do we distinguish between macro-driven price action – risk-on and risk off behavior tied to the global recovery – and sector-specific performance tied to underlying credit
trends and supply/demand technicals? While distressed credit markets often move in sympathy with equity markets, the correlation with rates is stronger. Fig 2a illustrates this with non-agency MBS Option ARM prices, which have moved in synch with 10-year swaps over the past year. Since distressed MBS
prices rise as rates sell off, and fall as rates rally, Option ARMs exhibit negative duration (that is, their prices moves oppositely to Treasury bonds). This unusual relationship occurs because rates tend to rise in response to economic optimism, and fall during risk-off crises.
Option ARM spreads are typically 600-800 bp, so correlated spread compression and widening is more important than the level of the swap benchmark. Although we do not show the charts, subprime MBS,
CMBS, mezzanine CLOs, and other distressed structured credit products exhibit analogous behavior (as long as dollar prices are low enough for spread variation to outweigh movements in benchmark rates).
Quantifying the relationship between rates and prices (via six-month empirical hedge ratios) allows us to decompose price action into macro-driven and sector-specific. For example, Option ARMs underperformed their rate/macro hedge in early summer, as weak housing numbers and Maiden Lane supply weighed on the mortgage market, but exhibited considerable resilience in August after S&P downgraded the US (Figure 2b).

If we use rates as a gauge for macro sentiment, we can attribute most of the 2011 price decline in Option ARMs to global risk aversion. Similar observations apply to other structured credit markets such as CMBS and CLOs, although technical factors were a relatively larger driver of subprime prices (approximately 1/3 of price performance, by our attribution). Is structured credit attractive, given continued macro uncertainties? In our illustrative case of Option ARMs, hedged performance this year has been approximately 3% plus carry. With typical carry on Option ARMs at 4-5%, this sector would be worth careful consideration even if expected hedged price performance were zero.

In fact, there are reasons to believe Option ARMs may be cheap to their macro hedge. Observed market betas imply rather draconian collateral deterioration if the economy stumbles, yet current-to-delinquent transition rates continued to improve over the past year even as home prices declined, growth softened,
and confidence weakened (the same is true for Alt-A and subprime mortgages). This suggests that performance might deteriorate only modestly in the event of a mild recession. If collateral performance continues to improve, price performance should eventually follow – even if the macro outlook remains
unchanged.

To take advantage of opportunities in Option ARMs or other products, how do we decide how much to hedge versus macro uncertainties? Generally, we can separate structured credit valuations into expected carry, richness or cheapness versus the macro/rates hedge, and overall view on macro
direction. Idiosyncratic risk to rates (eg, the Fed distorts the curve using “operation twist” reallocations), and to the specific sector (eg, mortgage put-backs or new government modification programs) can also change betas, requiring more asset-specific analysis on the precise hedge. However,
the macro prognosis applies to all sectors.

We believe the risks of a double dip have increased, but that the likelihood of a severe downturn remains low. If we view the recovery in the context of a banking crisis, we should expect prolonged weak growth, shocks to confidence, and sovereign debt turmoil. The existence of these phenomena does not in itself imply a second major contraction (although it may appear so when viewed in the context of an ordinary recession). Furthermore, many structured credit markets have already priced in a mild downturn. In this light, we believe the market is currently presenting a number of attractive investing opportunities.

Leave a Comment