Monitoring Market Trend With COT Metrics

In case you are not familiar with it, let's have a quick overview of the widely proclaimed and yet widely misunderstood Commitments of Traders (COT) report. The primary agency with regulatory supervision of commodity futures and options markets in the United States is the Commodity Futures Trading Commission (CFTC). The CFTC's stated mandate is to ''protect market users and the public from fraud, manipulation, and abusive practices related to the sale of commodity and financial futures and options, and to foster open, competitive, and financially sound futures and option markets".

In line with this mandate, the CFTC collects and circulates data on Open Interest (number of contracts held, long and short) for markets in which 20 or more traders hold positions equal to or above the reporting levels established by the CFTC. In practical terms, this means almost any liquid financial market publicly traded in the United States, including currencies. The reason for doing so is to nurture a level playing field, so that the price effects that could result from large swings in market participants' buying and selling activities can be known in a reasonably timely and open manner.

Now, the question is, why should we care about the futures markets since we are trading the cash market? This is because the futures market for currencies leads the cash market, and both markets actually trend in a parallel fashion. This means that, if we were to overlay the price plot for the EUR/USD Forex pair on top of the price plot for Euro futures, they would look pretty much the same. We can therefore assume that the currency futures price is a proxy for the Forex market. In simple terms, changes in buying and selling activity in the futures world would eventually affect Forex price movement.

The primary groups of traders traditionally covered by the COT report include the following:

• Commercials - Large corporate entities that use futures markets to hedge against business risks pertaining to the commodity which they manufacture or distribute (e.g. a grain pool which sells wheat on the open market). Commercial traders are typically counter-trend traders, not speculators.
• Large Traders - Financial market entities who speculate on the price movements of the underlying commodity without either providing or taking physical delivery of it (e.g. a hedge fund which trades and invests in various assets on behalf of its clients). Large Traders are typically trendfollowers.
• Small Traders - Primarily private traders holding positions in futures or options that are below the reporting threshold specified by the CFTC. Since Small Traders do not report to the CFTC, their positions are inferred as the residual of Commercials' and Large Traders' open interest in each market from the known total.

The COT report is published every Friday by the CFTC, based on reporting data submitted the Tuesday prior. COT data can show us the Net Long or Net Short positions taken by the above three categories of market participant, and highlight significant changes from one week to the next which may warn us in advance of accumulation/distribution campaigns that could affect price. The CFTC does not publish corresponding price data, but this critically important data can be obtained from other sources.

There are many different ways in which COT data can be interpreted. Some analysts look for extremes within a range of 6, 12, 24 or 36-month look-back periods by calculating a simple Stochastics index on the respective positions, often with the corresponding price series plotted as an overlay. This will tend to reveal when one category of trader hits a multi-period extreme of buying or selling activity (particularly at an apparent price high or low), which is thought to act as a warning of a potential price reversal.

While this method maybe perfectly sensible, I believe the best and simplest way to use COT data is to confirm a high level trend, and most importantly, changes in the trend. To confirm a high level trend, I do not look at the Commercial Traders' position data, but rather at the Large Traders'. Again, this group has a primary focus on trend following. As independent traders, isn't that exactly what we are trying to do as well?

In addition to the fact that Commercial Traders are counter-trend traders, studies have shown that Commercial Traders tend not to make money from futures trading, but rather to lose! Again, their primary interest is to hedge against risk in the markets in which they operate - not to speculate on price movement. Losses from futures market trading are therefore merely a cost of doing business for Commercials - just like buying insurance. In other words, go long when the Commercials are going long (or short when they are going short) and most of the time we will lose.

We want to trade with the trend, not against it. We want to pay attention to the group that is going to help with our trading, and it's usually not the Commercials! Therefore, my primary use of COT is simply to look at how Large Traders are positioned in relation to price action itself. I do not over burdened myself with look-back periods, Stochastics formula, or anything like these. Every Friday I obtain the latest COT positions data and corresponding price series, enter them into an Excel spreadsheet which then calculates the Net Position (i.e. long contracts minus short contracts) and then chart the respective series side-by-side.

If I see evidence of a high-level trend on price, and that Large Traders are on the same side of the market, I have reason to believe the trend is valid. Alternatively, if my price chart analysis shows that a high level reversal is setting up and that Large Traders have flipped from Net Short to Net Long (on a bottom), or Net Long to Net Short (on a top) consistent with the anticipated price reversal, then I have further reason to believe the reversal is actually happening. COT is therefore a high level trend confirmation tool, not usually a timing tool.

To accomplish the above objectives, I plot weekly Tuesday closing price on one chart panel, and concurrent net positions of Commercial versus Large Traders on the adjacent panel (bearing in mind that because the Commercials are always on the opposite side of the market from both Large and Small Speculators, the two plots will be perfectly symmetrical) as shown in the chart below.

In conjunction with standard trendline and Swing Point analysis, I then look for the following types of readings on the COT display:

Reading	Description
Bullish	Large Trader net positions line is above zero and rising: Net Long and following the uptrend.
Bullish Crossover	Large Trader net positions line crosses the central axis from below: changing bias from Net Short to Net Long, which may confirm a price bottom.
Positive Divergence	Large Trader net positions line makes a higher low in relation to a lower low on price: a price bottom (they are not following through to the downside).
Bearish	Large Trader net positions line is below zero and falling: Net Short and following the downtrend.
Bearish Crossover	Large Trader net positions line crosses the central axis from above: changing bias from Net Long to Net Short, which may confirm a price top.
Negative Divergence	Large Trader net positions line makes a lower high in relation to a higher high on price: a price top, (they are not following through to the upside).

It should be noted however that not all readings mean what they appear to mean, and not all actions of Large Traders can be assumed to be correct at all times. Thus, when Large Traders add to a net position but price thereafter does not penetrate an important level in line with trend, we can assume the undertaking was a failure, which could verify a technical analysis calling for a reversal of some kind. Failure signals can therefore be as useful as confirmation signals.

To some seasoned traders, the approach described above may seem to be too simple and hence questionable. However, the proof, as they say, is in the pudding. The sample COT chart for the US Dollar Index covering the period from January 2007 through December 2009 as shown above plots price versus Commercial and Large Trader net positions. I have labeled all crossovers, readings which are expected to confirm tops or bottoms based on price chart analysis undertaken separately. The results of these crossovers in relation to subsequent price action are summarized below:

• Reversal #1: Bearish crossover on Feb. 20th, 2007. Price on the USDX was 8410. Large Traders remained Net Short from that point through to Dec. 18th, 2007, when price had fallen to 7743. A short on the USDX using these two crossover signals to confirm the entry and subsequent cover long was worth (8410 -7743) = +667 points.
• Reversal #2: Bullish crossover on Dec. 18th, 2007. Price on the USDX was 7743. Large Traders went Net Short on an abortive move that ended up quickly resolving to the prior downtrend (an example of a failure), and thus their position reversed again on Dec. 31st, 2007, when price had actually fallen further, to 7670. The maximum loss on this failure signal was limited to (7670 - 7743) = -73 points.
• Reversal #3: Bearish crossover on Dec. 31st, 2007. Price on the USDX was 7670. Large Traders went Net Short again, and remained on that side of the market through to May 13th, 2008, when price had fallen to 7350. A short on the two crossover signals was worth up to (7670 - 7350) = +320 points.
• Reversal #4: Bullish crossover on May 13th, 2008. Price on the USDX was 7350. Large Traders flipped Net Long, and remained on that side of the market through both an interim top, which came Mar. 3rd, 2009 at a price of 8952, and beyond to the next crossover date of May 19th, 2009, when price had come down to 8215. To the highest high in March, the long was worth up to (8952 -7350) = +1602 points. To the May crossover date, the position was worth (8215 - 7350) = +865 points.
• Reversal #5: Bearish crossover on May 19th, 2009. Price on the USDX was 8215. Large Traders flipped short, and remained on that side of the market through to Nov. 24th, 2009, when price had come down to 7517. Using the crossover signals again to confirm an entry short and cover long yielded an opportunity worth (8215 -7517) = +698 points.

The above examples show that from February, 2007 through November, 2009, a straightforward analysis of Large Trader net position reversals on the US Dollar Index confirmed tradable opportunity in the range of 3,000 points. Note that this is not to suggest that you should approach COT data looking for extremely simplistic, 'black-box' trading signals; but rather, that you use the information to confirm other forms of analysis.