Seeking Alpha
Recommended for you:
Arbitrage, quantitative investing, strategy-based indices, systematic strategies
Profile| Send Message|
( followers)  

Summary

  • Low volatility indices have become popular.
  • Creating an Ultra-Low Volatility Index involves trade-offs.
  • I have some provisional ideas, but no definitive solutions yet.

Recently, I have been playing with ideas about how to create an Ultra-Low Volatility Index.

Provisionally, I am exploring the simple idea of adding an allocation to TMF (NYSEARCA:TMF), a 3X 20+ year Treasury ETF, and to SPLV (NYSEARCA:SPLV), the PowerShares S&P 500 Low Volatility ETF.

I have zeroed in on some simple rules. Admittedly, the testing period is far too short, so any results are provisional. I need more time, data, and real-world performance data.

Here are the provisional strategy's rules.

I. Go Long SPLV with 80% of the dollar value of the portfolio.
II. Go Long TMF with 20% of the dollar value of the portfolio.
III. Rebalance annually to maintain the 80%/20% dollar value split between the positions.

Here are the results in a linear scale:

(click to enlarge)

Very provisionally, I am intrigued by the lower drawdown, higher CAGR, Sharpe, and 0.33 correlation to the SPLV. But we do not have much data. However, it is heartening to see a pattern emerge of outperformance during tough years for the SPLV, and lagging performance during very bullish years--not always, but often, the hallmark of a conservative strategy.

In 2011, we really see the provisional Ultra-Low Volatility strategy shine in a choppy market:

(click to enlarge)

What I like about zeroing in on correlations during choppy markets, is that any seeming correlation from two portfolios both drifting upwards together during bull markets in equities is eliminated. Ideally, I like to see less correlation during period of market stress, as we do in 2011 with the -0.05 correlation of the strategy to SPLV. And in 2011, we get a dramatic boost to the Sharpe ratio.

Now, let us compare the provisional Ultra-Low Volatility Strategy to the S&P 500 (NYSEARCA:SPY) ETF:

(click to enlarge)

And in 2011:

(click to enlarge) Provisionally, still very impressive with even less correlation, and a bigger boost to the Sharpe ratio.

Thus far, my logic for exploring this potential strategy is that a leveraged ETF, because of the leverage inherent in the instrument, negates the need for even more dangerous portfolio margin debt, which is often inherent in risk parity strategies. And to avoid open-ended tail risk, and I am eating convexity in TMF as opposed to capturing the convexity inherent in shorting TMV (NYSEARCA:TMV).

In addition, my thought in using the S&P 500 Low Volatility ETF for one of the strategy's legs was to minimize the need for the bond allocation, which could be a serious drag in a rising interest rate environment. My reasoning is that low volatility stocks need less of a bond hedge (which may not be a hedge at all in certain interest rate environments).

As I assess this strategy, my goal is to compare the strategy's MAR (CAGR/Max Drawdown) to the MAR of the SPLV and SPY.

As of now, I have no firm conclusion, but I believe it is a wise allocation of my time to explore these instruments further given the provisional data. The strategy is not innovative like Structural Arbitrage, so I am worried about outperformance going forward.

Source: Some Ideas For An Ultra-Low Volatility Index