Fear Gauge Spikes: Let’s Play Hot Potato

For what risk does the commodity investor get paid? At what point is the fear gauge so high the risk gets passed like a hot potato? The answers to these questions will help explain why post the global financial crisis there has been a link between VIX spikes and commodity losses.

Let’s address the first question of what risk the commodity investor takes to be compensated. While there are five fundamental sources that drive the commodity asset class returns, the insurance risk premium is a major source of return and is the one behind the VIX – commodity link. Please see below for the chart of fundamental sources and components of commodity returns and click here to hear more about them in an educational slideshow.

5 Commodity Return Sources

The insurance risk premium is available to long-only commodity investors since there is a gap that needs to be filled between producers and commercial consumers that are hedging. Remember, the futures markets exist to facilitate hedging, not to forecast prices. The producers go short to protect against price drops and the consumers go long to protect against price increases. However, the producers need protection against price drops more than consumers need protection against price increases. The reason this is the case is supported by two economic theories: 1. Hicks’ theory of congenital weakness that argues it is easier for consumers to choose alternatives so they are less vulnerable to price increases than producers are to price drops, and 2. Keynes’ theory of “normal backwardation” that argues producers sell commodities in advance at a discount which causes downward price pressure, which converges to the spot at the time of delivery.  This results in net short hedging pressure from physical users as shown in the graph below:

Hedging Pressure

This leads us to the second question, “At what point is the fear gauge so high the risk needs to be passed like a hot potato?”

As published in a white paper by Cheng et al., while in normal times, or in other words “pre-crisis”, commodity investors accommodate the needs of commercial producers by providing the aforementioned insurance by taking the long side of the futures contracts that the commercial consumers forego. However, in times of financial distress, or “post-crisis”,  commodity investors reduce their net long positions in response to an increase in the risk as measured by VIX, causing the risk to flow back to the hedgers.

What is happening, as Tang and Xiong argue, is that although the risk sharing improves as the presence of commodity index traders increases, the expense of risk spillover from outside markets increases.  Simply stated, the risk for investors to provide insurance becomes too high to bear. Like the convection of a current of air that flows from a high-pressure area to a low-pressure area, the risk flows from the more distressed commodity index investors to the the less distressed hedgers – beginning the game of “hot potato” since no one wants to hold the risk. The result is a collapse in open interest as shown in Cheng’s paper in the table below.

VIX Commodity Open Interest

How does this translate into linking VIX spikes to commodity dips? Using weekly return data on a rolling basis daily back to January 1990, of a total of 215 times that VIX spiked more than 20%, 142 occurred prior to Sept 2008 and 73 occurred after Sept 2008. In the pre-crisis, of the 142 times that VIX spiked >20%, commodities fell 82 times or 58% of the time with an average return of 6 basis points during the weeks of VIX spikes. However, the post crisis number of weeks on a rolling basis where VIX spiked >20% was 73 and commodity returns were negative 65 of those times or 89% with an average return of -3.4%. Please see the charts below that demonstrate this:

Source: S&P Dow Jones Indices. Data from Jan 1990 to July 2013. Past performance is not an indication of future results. This chart reflects hypothetical historical performance. Please see the Performance Disclosure at the end of this document for more information regarding the inherent limitations associated with backtested performance

Source: S&P Dow Jones Indices. Data from Jan 1990 to July 2013. Past performance is not an indication of future results. This chart reflects hypothetical historical performance. Please see the Performance Disclosure at the end of this document for more information regarding the inherent limitations associated with backtested performance

A similar result holds when examining commodity returns for the week following a VIX spike >20%. Pre-crisis returns were 0.00% and post-crisis returns were -2.2%. Cheng’s paper finds a similar result for individual commodity futures markets via regression analysis as shown below:

We report coefficients from a weekly regression of commodity returns as the left-hand side variable on contemporaneous and one lag of changes in  the VIX as right hand side variables, controlling for lagged commodity returns, percentage changes in the BDI, changes in the Baa credit spread, and  changes in inflation compensation. Each row reports coefficients for a different commodity and each set of columns reports coefficients for different  sample periods. For brevity, only the coefficients on the contemporaneous change in VIX are reported. Coefficients are reported where both returns  and the VIX are in basis points. We use the Newey and West (1987) construction for standard errors with four lags. */**/*** denotes significant at  the 10%, 5%, and 1% levels, respectively.

We report coefficients from a weekly regression of commodity returns as the left-hand side variable on contemporaneous and one lag of changes in the VIX as right hand side variables, controlling for lagged commodity returns, percentage changes in the BDI, changes in the Baa credit spread, and changes in inflation compensation. Each row reports coefficients for a different commodity and each set of columns reports coefficients for different sample periods. For brevity, only the coefficients on the contemporaneous change in VIX are reported. Coefficients are reported where both returns and the VIX are in basis points. We use the Newey and West (1987) construction for standard errors with four lags. */**/*** denotes significant at the 10%, 5%, and 1% levels, respectively.

In conclusion, something to watch as an indicator of when the risk on – risk off environment might end this game of hot potato may be how the quantitative easing and policy drive the correlations between commodities and equities, which seem to have peaked, but not so obviously to declare game over yet.

Source: S&P Dow Jones Indices. Data from Dec 1994 to July 2013. Past performance is not an indication of future results. This chart reflects hypothetical historical performance. Please see the Performance Disclosure at the end of this document for more information regarding the inherent limitations associated with backtested performance

Source: S&P Dow Jones Indices. Data from Dec 1994 to July 2013. Past performance is not an indication of future results. This chart reflects hypothetical historical performance. Please see the Performance Disclosure at the end of this document for more information regarding the inherent limitations associated with backtested performance

The posts on this blog are opinions, not advice. Please read our disclaimers.

8 thoughts on “Fear Gauge Spikes: Let’s Play Hot Potato

  1. Dan

    Isn’t there a less complex explanation such as commodities and the S&P 500 have simply become highly correlated over the last five years and for an investor to gain a true non-correlated return he or she should look for actively managed commodity programs such as trend following so that they can take advantage of down moves as well as up moves with the added advantage of non-correlation to their exposure to equities.

    Reply
    1. Jodie GunzbergJodie Gunzberg Post author

      Commodities and the S&P 500 (and other asset classes as well) haven’t “simply” become correlated over the last five years but there are fundamental reasons behind why they have become more highly correlated. The supporting logic is that the unprecedented quantitative easing after the crisis has created a “Risk-on / Risk-off” environment so markets are pricing in the bimodal outcome – either policy works for a recovery or does not work and causes a more serious crisis. As Hilary Till points out in a “Commodities Now” article published in March 2013 and Cheng et al. points out in the paper referenced in the post, one aspect of the RORO environment post-crisis is that commodity investors in their own financial distress become liquidity consumers rather than providers so will not take the risk to provide the insurance premium to fill the hedging gap. In turn that causes the open interest to collapse because of the inability to take risk.

      The first thing to be clear about when discussing managed futures programs in your question is that we are talking about commodity-only CTAs and not programs that also include financial futures since they have different fundamentals. As I have discussed in two book chapters, Absolute Returns in Commodity (Natural Resource) Futures Investments in Hedge Fund & Investment Management (Edited by Izzy Nelken) and also in The Long and Short of Commodity Futures Index Investing in Intelligent Commodity Investing, there are two major opportunities to capture returns in commodities, which are cyclical opportunities and systematic opportunities. The blog post (and long-only commodity index investing) is discussing part of systematic opportunities which have to do with the economic function performed by commodity futures markets. However, trend following systems can capture cyclical opportunities in commodities because only price can respond to balance supply and demand. This is since commodities in the short-run cannot be drilled, mined and grown causing relatively slow cycles of inventory building, which in-turn drives patterns of commodity prices to trend up and down. So, when there is a supply/usage imbalance in a commodity market, its price trend may be persistent, which may be captured by trend-following programs.

      Reply
  2. Pingback: RECORD BREAKING MONTH: Commodities Love War and Heat | S&P Dow Jones Indices

  3. Pingback: Commodity Beta: Hogs-Wild? Hardly. Energy Fills the Thrill! | S&P Dow Jones Indices

  4. Pingback: Commodities Crystal Ball: What do the FUTURES hold? | S&P Dow Jones Indices

  5. Pingback: Taking Risk and Making Money | S&P Dow Jones Indices

  6. Pingback: HOT PORK ROLLS | S&P Dow Jones Indices

  7. Pingback: COMMODITY COMEBACK | S&P Dow Jones Indices

Leave a Comment

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>