On November 14th, while the S&P 500 (SPY) dropped 1.39%, an equal weighted composite of closed-end funds (including long equity, covered call, MLP, and high-yield bond), debt-buying companies and mortgage REITs fell by 3.89%. The worst performers in the group were Asset Acceptance Corporation (AACC), which fell 10.66% and the PIMCO High Income CEF (PHK), which fell 6.64%.
This accelerated a trend over the past month during which the closed-end fund sub-group fell 8.76% and the debt-buyer/mortgage REIT sub-group fell 18.86%, versus the 5.12% drop in the S&P 500.
Note: CEFs include PHK, EVV, BOE, CII, RVT, EOS, KYE, NHS; Debt-Buyers include AACC, PRAA, and ECPG; Mortgage REITs include NLY.
Weakness has occurred in two waves over the past month, with debt-buyers and mortgage REITs selling off in early November and the sell-off in Closed End Funds accelerating over the past week. It has also been echoed by weakness in other high-yielding sectors that are not as highly levered, like Master Limited Partnerships, proxied below by the chart for the Alerian MLP Index ETF (AMLP).
I believe that this effect may be due to deleveraging by hedge funds and other levered investors, which have engaged in carry trades with high-yielding CEFs, MLPs, high-yield bonds, currencies and other securities. The relative "crowdedness" of this trade became evident in August, when the PIMCO High Income CEF traded at a premium to its Net Asset Value of over 74%; PHK's current premium to NAV is 27%.
What are the Implications? If deleveraging by levered investors is beginning to occur, it raises the risk of negative feedback loop that could result in a sharper sell-off in all risk assets, rising correlations between assets, and rising liquidity risk premium and volatility.
The following chart shows the 1mo return difference between the leverage composite charted above and the S&P 500. The last month has seen the sharpest relative weakness for leverage since October 2011 and late 2008-early 2009.
In 2011, relative weakness for leverage on this magnitude was followed by a rebound in relative performance. In the 2007, 2008 and 2009 deleveraging episodes, it was followed by increasing volatility and bouts of relative weakness over the following several months. Indeed, as deleveraging accelerated in those instances, relative 1mo performance touched -10% on three occasions and -15% in late 2008.
One worrisome aspect for me of the current episode is that equity volatility as a whole, measured by the VIX, is still quite subdued. However, volatility may be starting to cluster upward, with the VIX currently trading above its 200d moving average.
Indeed, the term structure of VIX futures has recently switched from contango (rising term structure) to backwardation (inverted term structure). Inverted VIX term structures have previously been associated with future equity weakness and volatility clusters. Notice the flat price action of the VXX ETF over the past two months, rather than the persistent downtrend that stems from a steeply contangoed structure, and the bullish divergence in the RSI (see chart below). This type of price action is very similar to May-July 2011, prior to the August 2011 volatility cluster.
What to Do? If the deleveraging trend does not accelerate, we could see a technical bounce in the securities mentioned above (as happened in 2011). However, if it accelerates, here are a couple of suggestions.
First, what not to do. If volatility and the liquidity premium are on the rise, the following strategies should perform poorly:
- Carry Trades - Any trade that involves borrowing at a low yield and buying high-yielding securities (CEFs, currencies, high-yield bonds, etc) does poorly in a liquidity constrained environment.
- Volatility Selling - Volatility selling strategies, especially selling index puts, benefit from a gap between implied volatility and realized volatility (eg. realized turns out lower than implied) and from a less than perfect correlation between stocks in the index. During a liquidity/deleveraging episode, both realized volatility and correlations spike, causing losses in these strategies (which amount to selling catastrophe insurance).
Second, in this environment, defensive, low beta strategies - like the Russell low beta and low volatility ETFs (LBTA, LVOL, XLBT, XLVO) - high quality portfolios (eg. SDY or VDIGX) and trend following trading strategies have been shown to be good diversifiers.
For a short-term trade, I would also consider entering a long-position in VXX around $36/share (current price), with a stop 10% lower at $32.40 and a target of $54/share (based on past volatility episodes), for a roughly 5-to-1 risk-reward.