Micro and macro volatility in the oil market
– John Kemp is a Reuters market analyst. The views expressed are his own —
“Most probably we will continue to have reasonably high short-term volatility but in a narrower price range between $60-95 per barrel”.
That was the (accurate) forecast for crude oil prices given by Mercuria’s head of trading Daniel Jaeggi to the UN Commodities Forum in Geneva back in March [ID:nLDE62M0MT].
In fact front-month futures <CLc1> have been trapped in an even narrower range of $60-86 for the past 12 months, shrinking to $64-86 so far in 2010. Spot prices have barely budged since July last year, despite a substantial improvement in demand, as one puzzled investment bank noted recently.
Yet many traders complain high volatility is making either directional or technical strategies difficult to implement.
The apparent contradiction (high levels of very short-term price movement in a market trending sideways) highlights the different levels of volatility prevailing at different time horizons.
PICK YOUR HORIZON
Standard measures of volatility take price changes from the close of one business day to the next, averaged over the last 20 or 30 trading days, and then adjust them to an annualised rate.
On this basis, oil price volatility has been unusually low in H1 2010. Close-to-close volatility hit a 2.5 year trough of 19 percent in March.
Volatility increased as the rally broke down, peaking at almost 40.5 percent in late May and early June. But even that was low compared with peaks of more than 50 percent or 60 percent during previous periods of elevated volatility.
Close-to-close volatility is now back to just 28 percent, putting it in the 30th percentile of the distribution for all periods since the start of 2005.
But over other time horizons, volatility looks very different (http://graphics.thomsonreuters.com/ce/VOL-HRZN.pdf).
Intra-day volatility is higher compared with past periods. The real measure of intra-day volatility would take 5-minute or even 1-minute price changes over an hour or a day and annualise them. In practice that is too data-intensive for anyone not undertaking a full modelling exercise. But the daily trading range (high-low) is a reasonable if rough proxy.
Intra-day volatility (based on the high-low trading range) is currently running around 15 percent per year. While that is low compared with the close-to-close measure (28 percent) it is quite high compared with previous periods.
Intra-day volatility is currently in the 40th percentile (compared with the 30th for close-to-close volatility). Moreover, intra-day volatility peaked at the 90th percentile back in late May, compared with a peak in the 81st for daily volatility at the same time.
Traders who complain about large intra-day price swings are not imagining it; the market really is exhibiting unusually wide daily trading ranges at present.
Weekly volatility (measured from the close of business on one Friday to the next) is even more elevated compared with past periods. On a week-to-week basis, volatility is currently running at just over 41 percent, putting it in the 75th percentile of the distribution since the start of 2005.
TELLING A STORY
The challenge is to find a “narrative” explaining high levels of volatility at intra-day and weekly levels, yet comparatively low levels of volatility at the daily and yearly ones.
As former Federal Reserve Governor Lawrence Meyer remarked “When I’m asked what my profession was prior to joining the Board of Governors, I do not say that I was an economic forecaster, but rather a storyteller. As a forecaster, I had learned that neither my students nor my clients wanted to be buried in reams of computer output. They wanted me to tell a story that brought together in a coherent way the implications of the large numbers of economic indicators they saw as otherwise unconnected and therefore confusing”.
So can we attempt a “story” to explain different oil market volatility at different time horizons over the past 12 months?
Overall the market appears to have found a balance between bullish and bearish forces in recent weeks and months — which might mark an equilibrium (my view), an “uneasy truce” (the International Energy Agency’s view, expressed in its Medium Term Oil and Gas Markets Outlook) or a “state of disequilibrium” (Barclays Capital Commodity Refiner, Autumn 2010, page 62).
However this state is defined, it accounts for the broad sideways trend in prices over the past 9-12 months in a fairly narrow range of $70-80 or $65-85 per barrel, and the relatively low level of volatility over longer horizons.
But as the market has fluctuated between extremes or poles of euphoria and despair, prices show large week-to-week variations within the band. The price has “cycled” between the high end of the range and the low one, accounting for the high level of volatility at intermediate horizons such as the weekly level.
DAY TRADING AND HFT
Over shorter horizons, the market has been unusually stable at the daily level but much more variable at an intra-day level. One explanation focuses on increased presence of day traders and high-frequency or algorithmic traders in the market.
There is no reliable public data on how much of the daily trading volume in crude oil futures and options is accounted for by day traders or high-frequency traders (HFTs).
CME Group (which operates NYMEX) has conducted internal studies assessing the impact of HFT volume on volatility in various futures contracts, including oil.
The results were briefly cited in CME’s presentation to the Commodity Futures Trading Commission (CFTC)’s Technology Advisory Committee (TAC) last week, but have not been published. The CFTC itself does not appear to have reliable numbers on HFT let alone day trading volumes.
But HFT volumes already account for around 50 percent of all trading in some equity markets, and there is clear evidence the strategy is increasingly being applied to commodity markets. The point about HFT and day trading is that most positions are held for an extremely short term. In most cases, positions are not carried overnight and must be closed before the end of the session.
Could an increase in HFT and day trading be increasing intra-day price movements, while also dampening close-to-close movements, since HFT and day trades are normally reversed by the time the market closes?
CME Group insists HFT has dampened volatility by boosting liquidity, and cites unpublished internal studies to reinforce the point. But the answer is we don’t know because there is not sufficient information on HFT and day-trading volumes in the public domain. It might become clearer if the CFTC obliges exchanges to make more data available as part of the TAC review.
However, heightened activity among traders taking a very short term (sub-24 hours) view on the direction of oil prices and spreads, and using low-latency technology to arbitrage between oil and equity markets, would be consistent with the observed increase in intra-day price movement while close-to-close movement remains limited.