Interdum stultus opportuna loquitur...

Thursday, July 07, 2005

OilRant: How To Do Syncretic Analysis... Part 1

Note - from June 24th 2009, this blog has migrated from Blogger to a self-hosted version. Click here to go straight there.

Crude Oil is likely to make nuffnuff traders taste their own blood, and relatively soon. This also serves as a good illustration of how best to identify a likely short-term reversal in moderately-real time, and exploit it (although if you're doing indices, you will also need to include some analytics related to intraday TICK extremes - of which more at a later date).

First, why I think that we are near a top in Crude - based on what I refer to as a 'syncretic' approach (i.e., one that draws together fundamental, sentiment and technical analyses).


Fundamentals for Crude are still as bullish as ever they were - which is why my longer-term 'attractor' for the Crude price is still above $100; global reserves are overstated by as much as 50%, mostly due to political pressure on OPEC members to claim higher reserves in order to have a greater proportionate allocation of pumping rights. Demand is growing without obvious bound - even with rising prices, demand for Crude continues to grow faster than underlying 'steady state' economic growth (defined as growth in population plus growth in productivity; that totals about 2.7%).

The only possible fly in the Fundamental ointment is the abiotic oil hypothesis, which I find hard to ignore. If true, it also may be a partial explanator of the increased reserves outlined above; reserves could be replenishing through abiotic percolation rather than simply being overstated by corrupt bureaucrats (although the Transom Corollary to Occam's Razor indicates that any explanation that involves corrupt bureaucrats is invariably the correct one).

Crude reserves are just one link in the supply chain - there is also extraction capacity and refinery capacity to consider, although refinery capacity is more interesting for the pricing of final products (fuels and lubricants, as well as feedstocks for the plastics industry); refinery capacity is currently running at over 90% and new refining capacity represents investment with long lead-times (so even if the price of crude was to fall, supply pressures would remain in end-product markets for some time).

Extraction capacity is even slower to come on-line, since it usually requires exploration spending before extraction can be endeavoured (most proved fields are already tapped to some extent, with the exception of modest-sized reserves in the ANWR).  There has been a great deal of damage to extraction capacity as a result of the invasion of Iraq, and any flare-up towards Iran would almost certainly result in massive (deliberate) damage to oil loading facilities, and/or increased insurgent targeting of the Straits of Hormuz - the worlds largest oil-transit chokepoint.

As such, we can't look to supply to help out in the next few months. A good deal of demand is generated as a by-product of other economic activity (i.e., transportation) and so that side of the market can't adjust readily without a widespread economic contraction (which I think is going to happen anyhow, but it's not going to be tradeable in the short term).


The primary measure of sentiment in the oil market is the positioning of the various trader categories in the Commitment of Traders reports published by the Commodity Futures Trading Commission (CFTC).

Of particular interest to me, are two numbers - based solely on the idea that markets attract new, dumb money on the wrong side at price extremes. The two numbers of interest are

  • The ratio of long positions to short positions for 'Non-Reportable" traders; and
  • the share of Non-Reportable Long Positions in total Open Interest.

I call the first ratio the "Dumb Bull Ratio" or DBR; it works best in markets where the Non-Reportable traders control a decent chunk of total open interest (OI) - for example the S&P500, where the share of Non-Reportables in total OI is about 40%. Since the Crude Market contains far fewer small-money traders than my favourite markets (total non-reportable positions are less than 20% of OI), it is somewhat less reliable.

Still, it is worth using, particularly when the news if full of blather about the commodity in question.

On the charts below, I've highlighted the period in which oil prices were in the news almost daily. It also happens to overlap with the first short-term short I advocated in Oil, from just above $58 - which paid handsomely prior to a reversal back up. A couple of things bear mentioning, which are not shown - the main reason that DB%0I failed to make a higher high on the second price peak, is because short interest from Commercials (not shown) rose, pushing up overall  Open Interest.

Crude with DB%OI

At the same time, notice that the Dumb Bull Ratio (below) was climbing as the market made its second peak, and actually peaked after the price peak, indicating that small trades continued to believe that the price decline was temporary - note how the DBR similarly continued to decline after prices had already turned back up. that's why I call it the Dumb Bull Ratio.

Crude with DBR

At present, Dumb Bulls are still a relatively small slice of overall open interest (at just under 9%), but Non-Reportable long positions will almost certainly rise this week since the market has reversed hard and made a new high. I also expect this week's CoT report for Crude to show that Commercials have built short positions.

The Dumb Bull ratio is in the top third of its historical range (the chart is hard to read, but what do you want for free?), and again, its reading will rise when the latest batch of CoT reports are released this Friday.


The technical picture is pretty unambiguous - less so than the Sentiment picture, which almost always requires extrapolation at or near swing tops.

I always do technical analysis over multiple timeframes; here's a broad outline as to how.

A Weekly chart defines the longer-term trend. Its moving average it to be respected, and counter-trend positions are assumed to be temporary until proved otherwise. In an uptrend, and overbought oscillator reading shows a high-probability place to change to short short-term (daily and intraday) bias, but only when coupled with a divergence on a secondary oscillator (as shown below, in October and March/April). When the 'core' oscillator pulls back to oversold, the reversal back in line with the dominant trend  doesn't require a confirming divergence (as in December and May).

At present, the weekly chart of Crude Oil shows two important things; the chart is overbought as per the Wiliams %R (always my primary oscillator), and the CCI has made a lower high than the March-April high. (The CCI is almost always my secondary oscillator). In a way, that's more important than whether or not there is a divergence like those shown in October and March/April; this indicates a higher price high than April, yet a lower CCI high - in other words, a longer-framed divergence. the fact that the CCI in April was higher than the CCI in October indicated strongly that the pullback was not one that should signal a genuine medium-term trend reversal; this time it may be different.

Notice how the 50-week moving average gets a lot of respect as a rising support - that gives some idea as to the target range for a pullback.

Crude Weekly

OK: that's step one. We have a possible significant reversal on the hands based on the weekly.

From there, move to a daily chart. Notice the divergences again (they are a little more timely on Daily Charts than on weekly) and notice how the 200-day moving average is respected as rising support.

Crude Daily

that gives another heads-up (really just a higher-resolution confirmation of the weekly).

From there, I look to hourly charts to refine entries. I'm not going to stick in an hourly chart here, because I can't be bothered - but hourly charts finished overbought yesterday, and will likely open even more overbought today... and there's a divergence on the intraday CCI, too.

Suffice it to say that an intraday overbought during the day session will be the highest-probability short entry for some time: last week's the recent spike high at $60.95 a week or so ago, prevented a divergence on daily charts that would have made for a nice VERY short-term intraday short signal - it meant missing a $4 pullback, but you can't trade a signal that 'nearly' exists... that's a good way to go broke.