Then: Risk-On/Risk-Off. JPMorgan’s Marko Kolanovic head of Equity Derivatives Strategy explains:
Over the last 5 years, Treasuries and Equities had strong negative correlation. This was the risk-on / risk-off (RORO regime) in which Treasuries were the most broadly used ‘safe haven’ asset. In the RORO regime, investors would hold treasuries and sell them to buy risky assets (and vice versa) while being reassured that Fed will keep the price of Treasuries supported. While we are still on average in a RORO regime, the bond-equity correlation started significantly weakening due to increased risk of Fed tapering and a bond selloff. The effect of the Fed reducing the stimulus could result in lower bond prices as well as lower prices of stocks, commodities and other risky assets whose prices were inflated by the Fed’s stimulus.
Over the past month, in several instances bonds and stocks moved together as investors re-assessed the probability of early tapering. Figure 1 below shows equity-bond correlation (calculated from high frequency intraday data). Correlation turned positive on May 9, 22, and 31 and most recently over the past few days. May 9th and 31st brought better than expected macro data (jobless claims, consumer confidence and Chicago PMI). Ironically, positive data caused equities and bonds to trade lower on increased probability of tapering (good data were bad for stocks). Similarly, on May 22nd, bonds and stocks sold off as Bernanke indicated the possibility of tapering over the next few meetings.
And Now: the “Fed Regime“
A byproduct of these new bond-equity dynamics is that USD is losing its status as a ‘risk off’ currency. As expectations of more (less) stimulus pushes up (down) treasuries and US stocks (both USD denominated), resulting currency flows are weakening the historical negative USD-Equity correlation. Historically, USD had strong negative correlation to equities (i.e. EUR and EM currencies had a positive correlation to equities). This recent relationship is now undermined as treasuries are losing their appeal as a safety asset. This weakening of EM FX and EUR correlation to the S&P 500 (Figure 2) was also helped by investors putting money in US stocks, while avoiding European and Emerging markets in the last leg of market rally.
Fears of Fed tapering the massive QE program is now changing bond-equity correlation from a RORO regime towards a ‘Fed Model’ regime (coincidentally, the name ‘Fed Model’ was crafted in 90s long before invention of quantitative easing). We do not think equity-rate correlation will fully revert back to the ‘Fed Model’ regime, but the recent spikes in rate-equity correlation are worrying signs. Recent bouts of positive correlation of equities, bonds and commodities, suggest that the Fed’s stimulus inflated prices of a broad range of financial assets, and removal of the stimulus could create a tail event in which prices of all assets could go down. While it is expected that the Fed will try to avoid such a scenario by maintaining an appropriate level of stimulus, in the absence of more robust growth, this may turn out to be a difficult task (akin to driving a car without brakes). On this account, we expect more volatility in H2 as compared to the first part of the year. To reduce risk of a bond and equity tail event, investors could diversify ‘safe haven’ assets away from treasuries and into other assets that are at lower risk in case of tapering. For instance investors could increase allocations to cash or Equity Put options.
Helpful. It also appears that Marko and Tom Lee, who sees nothing but smooth sailing from here until S&P 2,000, don’t talk much.
And just so the message of JPM’s derivatives group is clear, they look at the unprecedented (and previously documented) surge in NYSE margin debt, which has risen at the fastest pace ever so far in 2013, and analyze the empirical evidence of what happens after such a radid move. The result is below:
Last month, NYSE published April data on aggregate debt balances in stock margin accounts. This measure shows how much funds were borrowed to purchase securities, and it reached all time high of $384bn. Net margin debt (calculated as a difference of debt in margin accounts and all credit balances) also reached a high level of $106bn, and the pace of net margin debt increase YTD ($87bn) was the highest on record. We have been asked whether this increase in leverage is a sign of an impending market selloff. To analyze the relationship between S&P 500 prices and margin debt we look at their historical levels over the last 15 years. Figure 7 shows a strong correlation between S&P 500 and NYSE net margin debit. Positive correlation between the S&P 500 and net margin debt indicates that clients tend to finance a fraction of their equity exposure with margin debt. We also note that peaks in margin debt are usually followed with a sharp market correction. However, this on its own does not imply that high margin debt leads to market correction (given the positive correlation of net margin debt and S&P 500, highs in margin debt coincide with highs in S&P 500).
To test for a causal relationship we looked at the changes in net margin debt against S&P 500 performance 3, 6 and 9 months afterwards. Figure 8 shows that large increases of net margin debt are indeed on average followed by weak equity returns. Note that the YTD increase of margin debt is the highest on record, as indicated by the arrow.
Another test we performed is to look at levels of margin debt normalized by the level of the S&P 500. Dividing margin debt by the level of the S&P 500 may give a more accurate measure of leverage (by remove the bias coming from correlation of S&P 500 and margin debt levels).
Figure 9 shows the ratio of margin debt to S&P 500 (red) as well as ratio of net margin debt to S&P 500 (blue). One can see that these normalized measures of leverage peaked prior to the tech bubble burst, in H2 2007 and H1 2008, and in H1 of 2011 – in all cases ahead of significant market corrections. While these are effectively only three data points and hence do not amount to a reliable statistical sample, we think that the quick increase of net margin debt, and high ratio of margin debt to S&P 500 do point to an increased probability of a market correction and volatility increase in the second half of the year.
But don’t worry. The Fed is on top of it. All of it.