## VIX Portfolio Hedging (VXH) Strategy 2012 Performance Review

Fri, Feb 1, 2013 | Jared Woodard

A successful portfolio hedging strategy does two things: it protects against market declines, and it imposes minimal costs in rising markets.

In a year like 2012, the ability to minimize hedging costs actually matters more than downside protection, and the VIX Portfolio Hedging (VXH) Strategy bested its benchmarks and peers by being more cost-effective.

Let’s start with a look at how the S&P 500 and the iPath VXX ETN fared in 2012. The histograms at fig. 1 show daily returns for each asset, with normal distributions fitted for reference. In short, there wasn’t much in the way of day-to-day risk to hedge.

*Fig. 1. SPX and VXX 2012 daily returns. Source: CBOE, Condor Options*

The peaked nature of those SPX returns (leptokurtosis) is intuitively consistent with the quiet year we saw: far more quiet days around 0% than big swings in either direction, and more quiet days than we would expect to see in a normally distributed series. The worst single-day return last year was -2.49%, the biggest gain was 2.46%, and the median return was 0.00019%.

Long volatility products tend to have negative mean returns and positive skewness because, most of the time, investors pay a risk premium for the protection that puts provide. The steeper term structure in recent years has added to that cost: in the case of VXX, most of the time the fund is constantly selling cheaper front-month contracts to buy more expensive second-month contracts. You can see how these features affected the VXX distribution for 2012. The worst daily return was -11.65%, the maximum gain was 10.53%, and the median return was -0.0087%.

*Fig. 2. S&P 500 2012 Drawdown. Source: CBOE, Condor Options*

It’s not just that daily return risk was muted, either: drawdowns were mostly shallow and slow. Fig. 2 shows the drawdown time series for SPX last year. The worst peak-to-valley drop was -10% and it took a full month (May). The maximum one-month historical volatility was just over 20%. All of this is to show that, in 2012, it was more important for hedging strategies to minimize costs than to keep heavy protection constantly in place.

Now let’s examine the signal history for the VXH strategy. For readers who aren’t familiar with the strategy, VXH allocates 1-100% of hedging capital to long VIX futures or VXX positions based on the strategy signal. The amount of hedging capital that should be made available depends on the composition of the portfolio being hedged, but for generic SPX-like equity portfolios we’re assuming up to 50% of the notional portfolio value is available if needed. The maximum signal level in 2012 was 11%, but the allocation level was near the absolute minimum threshold for much of the year: the median daily signal was just 2%.

*Fig. 3. VXH allocation signal (lower) and SPX (upper). Source: CBOE, Condor Options*

Here are two ways that VXH is different from other volatility-based strategies:

- Dynamic allocation: VXH changes hedging allocations very quickly when needed. Allocations are updated daily, and per fig. 3 there can be many weeks when no change is indicated at all; but when the market changes its tone, the strategy is quick to respond.
- Asymmetric comedown: Some strategies treat volatility spikes symmetrically, but as the VXX distribution above shows, the returns to volatility products tend to be heavily skewed, so it makes sense to withdraw hedging capital at a faster rate than it is committed. VXH reverts to a lower allocation more quickly than it ratchets exposure higher in order to match the asymmetric nature of equity volatility.

*Fig. 4. SPX, SPX + VXH, and SPX + 5% VXX portfolio returns, 2012. Source: CBOE, Condor Options*

An unhedged SPX portfolio (fig. 4, black) returned 11%, excluding dividends. The same portfolio hedged with VXH (green) returned 8.9%. A fixed allocation to long volatility products imposes greater costs than is necessary. Even a relatively small fixed 5% VXX allocation (red) had a considerable drag in 2012, returning just 3.3%. This is why we strongly advocate tactical, dynamic hedging strategies instead of fixed portfolio allocations. Volatility may be an asset class, but long volatility products are for trading, not for buying and holding.

*Fig. 5. VXH, VQT, VIXH, and HUS.U returns, 2012. Source: CBOE, Horizons, Condor Options*

Several interesting volatility-based exchange-traded products (ETPs) have been launched in recent years. Fig. 5 compares the performance of three VXH peers. 2012 returns are listed after each ticker symbol.

- VQT (1.8%): the Barclays S&P VEQTOR ETN consists of SPX exposure with tactical VIX futures positions (matching VXX weights), a 2% daily stop loss rule, and a 0.95% annual expense ratio.
- VIXH (-2.5%): the First Trust CBOE S&P 500 VIX Tail Hedge Fund consists of SPX exposure with semi-tactical purchases of VIX call options. The expense ratio is 0.60%.
- HUS.U (3.5%): the Horizons Universa US Black Swan ETF is a product listed in Toronto that consists of SPY exposure and options on S&P 500 products. As of December 31, 2012, the HUS.U portfolio was invested 94% in the SPY ETF, with a February 1000/500 SPX long put vertical financed by the sale of February 130 SPY calls. There is a 0.95% management fee plus a 20% performance fee.

VXH (8.9%) handily outperformed all of these alternatives. Returns for VIXH and HUS.U are higher in fig. 5 as they are plotted against contemporaneous VXH performance to make a direct comparison easier. VXH is available by subscription or in an individually-managed account.

Tags: HUS.U, spx, spy, VIX, vix futures, VIX Portfolio Hedging, VIXH, vxx

February 4th, 2013 at 3:45 am

I realize backtests aren’t worth much in the real world, but what i’m wondering is how this strategy would have done historically given different annual returns of the SPX itself. Obviously you’ve done a great job minimizing hedging drag to what was a great year for the SPX.

If the SPX was down 10% in 2013 for example, based on historical backtests, what is your expectation for VXH’s returns? Would the boost from the volatility returns be able to overcome the poor SPX performance? Or would you be expecting to reduce but not eliminate the 10% negative returns?

February 4th, 2013 at 9:56 am

Hi, there is more historical data and backtest information available in prior VXH posts:

http://condoroptions.com/category/blog/volatility-blog/vix-portfolio-hedging-blog/

The expected VXH return with SPX down 10% is path-dependent, since the strategy is designed to hedge against volatility in particular, not just negatively signed returns. E.g. in a 10% decline over the course of a week or two, VXH might cover the SPX return and then some, esp. if that decline occurs from a base of moderate or high volatility; a daily decline of 0.25% until -10% is reached would, by definition, not be very volatile and so would likely not be hedged away.