A nice bit of deviation for oil investors

That the lights have gone out at Bulb, so far the biggest UK energy supplier to blow its fuse board, can only reinforce our present obsession with energy, or – as looks likely this winter – the lack of it. And that reminds us that fossil fuels – reliable, efficient and running on proven, embedded infrastructure – have a future.

Indeed, there was a delicious irony to note early this month as the COP26 climate conference got under way in Glasgow. Minimal wind power meant power generator Drax had to fire up its coal-fuelled boilers to keep the lights on in Britain and wholesale prices within the UK’s electricity-transmission system peaked at £4,000 per megawatt hour, about 100 times their usual rate.

All of which indicates that there may well be profitable ways for retail investors to ride the energy transition (if that’s not too glib a term for a process that could easily take the next 50 years). One such way – and an easy one for investors to access via specialist exchange traded funds – is to take long-term bets on the oil price. For example, happy is the investor who was smart enough – or lucky enough – to buy oil for around $20 per barrel in the spring of 2020. By early this month, the price had quadrupled.

Using the spot price of North Sea Brent crude, the chart illustrates a generic way of identifying buy and sell opportunities. It could hardly be simpler. It takes a long-run time series of data – in this case, the weekly price of Brent crude starting in December 2002 – and uses a spread sheet to calculate the average price over the period, then the average price plus and minus one standard deviation, which measures the dispersion of values around their average. The investment strategy follows intuitively from that: to buy when the price falls below the average minus one standard deviation and to sell when it rises above the average plus a standard deviation.

Simply looking at the chart above indicates that for Brent crude the plan would have produced good results, certainly when buying at the average minus one standard deviation, which, in practical terms, meant buying below $43. Table 1 quantifies this. Buy oil below $40 and an investor would be hard-put to register annualised gains below 20 per cent over the periods shown in the table; and annualised gains of 40 per cent-plus are common. True, the caveat is that it would be difficult to deal at the table’s prices, which is partly because they are based on weekly returns. This explains why the price for mid-March 2020 – $32.25 – is so far below $40. It was the week the world woke up to the reality of Covid-19 and Brent crude’s price dropped 30 per cent on the week. The game also requires patience. Just five opportunities have come along in the past 18 years.

Relaxing the criterion and taking $50 as the buy threshold noticeably compromises the results. Although returns remain satisfactory, the variation between them becomes considerable. Of the seven opportunities to buy below $50 over the whole period, the gap between the best and worst six-month performance is 65 percentage points – from a 31 per cent gain for those who bought in November 2008 to a 34 per cent loss from dealing in August 2015. That said, it is encouraging that every buying opportunity shows a profit after two years. It reinforces the notion that, for crude oil at any rate, it would be hard to buy using the ‘minus one standard deviation rule’ and lose money.

However, the signal to sell at the average price plus one standard deviation, which works out at almost $97, is much fuzzier. That’s because, by definition, averages always change and it is debatable whether in future the oil price will be driven to $100-plus levels often enough to prevent the average price – currently almost $70 – slipping lower and lower. True, the law of unintended consequences may help investors. That is, the demonisation of Big Oil in the west may yet lead to such a shortage of capacity that the price of crude will be driven to levels that brings rejoicing in downtown Riyadh. We shall see.

Meanwhile, Table 1 shows that selling oil above $90 and even above $80 has, more often than not, proved to be a good move so far this century. Because the sale of an investment means there are no longer gains or losses to accrue, all that matters is whether future profits have been missed. Price momentum means there is a mildly odds-on chance that will be so after six months, whether the sale of crude was at $80 or even at $90. That is much the same as the effect of selling when, say, a stop-loss is triggered; the short-term outcome of such rules-driven moves is always a bit arbitrary. However, as the table shows, one or two years out it is more than likely a sale will have been a good move, presumably because over the longer period economic fundamentals will be reasserted.

As noted a few paragraphs ago, using long-run price averages and the variation around them is a useful generic way of spotting price anomalies in all sorts of investments. Basic statistics tell us that any time series of prices will show two-thirds of its data within the range of plus and minus one standard deviation of the average. So it follows that just one-sixth of the data will be at the extremes on either side of the one-standard-deviation frontiers. Yet extremes have a habit of reverting to their average values – hence the trading opportunity.

Sure, this assumes that price returns show a so-called ‘normal distribution’ around their average yet, at the margins, the returns of financial assets are not normally distributed. Feedback effects mean there is extra action at the extremes. Even so, most of the time the one-standard-deviation rule provides food for thought. It works best where price movements are likely to be cyclical, especially where the upwards drift caused by inflation can be factored out. That makes commodities especially suitable.

For equities, the rule works better in the aggregate – ie, for stock market indices – than for individual stocks and it should be applied to a measure of value than directly to prices. So, for example, the dividend yield on the FTSE All-Share index deflated by the UK’s inflation rate has been a useful indicator of the market’s cheapness and dearness. Where the rule is least likely to work is where the future is the haziest and where feedback effects – investment contagion, if you like – is most pronounced. In other words, don’t bother with it for technology stocks high on hope and thin on revenues let alone earnings.

As to ways into tracking the oil price, the London market offers plenty of exchange traded products (ETPs), many of which are provided by WisdomTree, which specialises in currency and commodity trackers. Table 2 provides very basic details of five. Although there are many others, they share similar aims – either to track an oil-price index as closely as possible, or to leverage up an index’s returns using daily compounding, or – in effect – to short sell an index, either with or without leverage.

Since oil is at a level from where its price looks to have more downside than upside, let’s quickly focus on the short-selling ETPs. Even without leverage, they are inherently risky essentially because they re-base their returns every day. To understand the effects of daily re-basing, consider what happens to the price of an ETP if it starts at 100, falls 5 per cent on one day and rises 5 per cent the next. Instinct might say it’s back where it started. Instinct would be wrong. The price would be 99.75 and the same would apply if the timing of the rise and fall was swapped. If that rings an alarm, consider what happens when daily returns are magnified three times. Small wonder WisdomTree’s three-times leveraged short Brent crude tracker (3BSR) has lost 95 per cent of its value in the past year. Take note.

bearbull@ft.com