An Alternative Passive Investment Strategy

21Feb13

Keeping with my theme from my last post, I’ll look at a common passive investment strategy, a 50/50% investment in SPY (S&P 500) and IEF (10-year UST), and replace the equity exposure with a (volatility-equivalent) investment in the volatility ETF XIV.

First our baseline, a 50/50% investment in SPY/IEF rebalanced monthly (frictionless):

20130221.01
[logarithmically-scaled, growth of $1]

Our alternative strategy is as follows:

When front month VIX futures will close below the second month (contango), go long at the close X% XIV + 50% IEF.

XIV is of course (much) more volatile than SPY, so to compensate, only buy the amount of XIV that is the volatility-equivalent of a 50% SPY position (based on the standard deviation of daily SPY and XIV returns over the previous 63-days).

Today, for example, that would be about 9% XIV, 50% IEF, and the remainder in cash.

When front month VIX futures will close above the second month (backwardation), go long at the close 50% SPY + 50% IEF (i.e. the original passive strategy).

Rebalance positions at the end of each month.

And the results, the baseline in grey and the alternative strategy in red (frictionless):

20130221.02
[logarithmically-scaled, growth of $1]

Numbers for the number lovers…

20130221.03

These two strategies are well correlated and very similar in terms of volatility and beta exposure.

The extra juice on the alternative strategy is mostly from the positive impact of the term-structure when futures are contangoed.

This alternative strategy isn’t without its own unique risks.

Some immediate ones that come to mind are (a) losses in volatility ETFs can accelerate very quickly during market crashes, and (b) should the VIX futures term-structure flatten or grow unstable (i.e. if today doesn’t predict tomorrow), the trade breaks down.

Remember, these are still relatively new products and we should be cautious about putting too much faith in a limited past to project the future.

P.S. a tangentially-related topic is the limited upside remaining in Treasuries (which I ranted about here, here, here, and here). That is of course going to be a drag on the strategies presented here, and yet another reason to be thinking about boosting returns even for simple passive portfolios like this one.

HappyTrading,
ms

. . . . .

To stay up to date with what’s happening at the MarketSci Blog, we recommend subscribing to our RSS Feed or Email Feed.



20 Responses to “An Alternative Passive Investment Strategy”

  1. 1 Sn

    Could you please elaborate on when there has been contango and when there has been backwardation for your calculation and how large the average time span is?

    • 2 MarketSci

      Hello SN – for this post, I’m measuring contango/backwardation simply based on when the front month is below/above second month VIX futures. That’s not how I do it in my own trading, but it’s good enough I think for this conceptual post.

      You can find historical VIX futures data at:
      http://cfe.cboe.com/Products/historicalVIX.aspx

      VIX futures have historically been contangoed about 82% of the time and in backwardation about 18% of the time. A switch has occurred on average about once every 16 trading days, with multiple long periods of up to 273 trading days without a switch.

      michael

  2. 3 John French

    Michael, have you got a excel file with this model that you could post please? No worries whatsoever if you haven’t. Do you know what the correlation of the alternative strategy is to the SP500?

    Thanks

    • 4 MarketSci

      Hello John – unfortunately no, but if you have trouble recreating these results, let me know and I’ll do my best to assist.

      Monthly correlation with SPY = 82%.

      michael

  3. I’ve actually been following a very similar strategy for some time. The added benefit (which I think is HUGE) is all the freed up cash this provides to allocate to additional non-correlated assets / strategies. Nice work.

    • 6 MarketSci

      Hello Jake – I didn’t know your were in to these vol ETFs. Good to have you around sir. I haven’t seen anything from you in a long time – is the blog taking a breather? michael

  4. 9 John Fischer

    Terrific stuff! “Risk-on” strategies which don’t involve buying overpriced stocks or other highly-correlated asset classes are a gift to your readers. However, I think even bond visionaries like Gary Shilling see Treasuries as a trade only at this point. Leads me to wonder what the comparison would be if you either zeroed out the IEF component, or scaled it over time in some way for expected return (e.g., just using the 30-day SEC yield).

  5. 11 Raphael

    Jake said “The added benefit (which I think is HUGE) is all the freed up cash this provides to allocate to additional non-correlated assets”

    Do you think allocating the remainder not invested in XIV in GOLD (GLD) or Real Estate (VNQ) instead of cash would be of any benefit?

    • 12 MarketSci

      Hello Raphael – I personally wouldn’t choose RE b/c of the high correlation with equities/XIV.

      I think GLD is a better choice – I was curious what those numbers would look like too, so I’ll put it on the to do list to show the numbers in a future post. michael

  6. Doesn’t the use of treasuries and your historic returns only look good for the long-standing bull market? Why wouldn’t this start posting losses when rates start to increase?

    • 14 MarketSci

      Perry – as I mentioned in the post and linked to in the post, I’ve written extensively about the limited upside remaining in UST and similar assets, and the inevitable negative impact on longer-term strategies like this one.

      The point of this post wasn’t to say “follow this strategy and get these great returns”. It was “here’s the historical benefit of the VIX futures term-structure over standard equity exposure in a common passive strategy”.

      But to answer your question: in a 50/50 split (or the vol-equivalent of a 50/50 split when using XIV), the equity/vol exposure is so much more volatile than the UST exposure, that I would still expect the strategy to be marginally profitable even when rates rose. That doesn’t mean it’s a good strategy, just that statistically speaking, in a 50/50 portfolio, UST is doing less than equities to drive returns.

      michael

  7. Why not just buy XIV and rebalance UVXY calls monthly? I day and swing trade XIV, not enough courage to catch overnight out of the blue without some protection.

    • 16 MarketSci

      This post is about an alternative to the vanilla 50/50 (or 60/40, whatever) stock/bond portfolio, so you’re strategy is probably super duper, it’s just not related to the topic at hand. michael

  8. 18 eber terandst

    Most likely I am missing something here, but . . .
    When in backwardation the expectation is for bad SPY results, so wouldn’t it make more sense to just directly to cash or bonds ?
    Thanks
    eber

    • 19 MarketSci

      Hello Eber – good question – backwardation isn’t necessarily indicative of poor SPY returns, only poor XIV returns (b/c of the daily impact of first/second month VIX futures converging upwards towards the actual VIX).

      It is true that, based on the very limited data we have available, backwardation has been mostly followed by poor SPY returns as well, but a lot of this is the result of 2008/09 and I think it’s too early to be confident using the VIX futures term-structure to time the actual market. I think there are other, much more thoroughly tested indicators for that.

      This is all off-the-cuff as I haven’t given this idea too much thought before. Will have to run more numbers… michael

    • 20 MarketSci

      Hello Eber – your question prompted this post (thanks for that):

      http://marketsci.wordpress.com/2013/02/26/stock-market-returns-based-on-the-state-of-the-vix-futures-term-structure/

      In short, my initial response was correct, but the question definitely deserved some further fleshing out. — michael


Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s