Tom McClellan Discovers a Strange Correlation
Here is a link to an article posted by Tom McClellan last year, which one should probably read first.
A brief summary: while searching for inter-market correlations that could be useful for traders, McClellan found out a few years ago that the positioning of commercial traders in euro-dollar futures appears to foretell the future direction of the stock market. The two charts (the S&P 500 index and the euro-dollar net commercial position) align remarkably well when commercial euro-dollar positioning is shifted forward by 52 weeks.
It would be pretty neat if it continues to work, as one basically gets a stock market road map a year in advance!
But why would it work? McClellan reasons in his article that since the euro-dollar contract is reflecting interest rates on dollars deposited in the European banking system and usually tracks LIBOR, one can assume that:
”… what we are seeing in their futures trading are responses to immediate banking liquidity conditions, and those actions give us a glimpse of future liquidity conditions for the stock market. These liquidity conditions are revealed first in the banking system, and then the liquidity waves travel through the stock market a year later.“
He then however adds this caveat:
“But even if we cannot identify exactly what makes something work, after a few years of seeing that it does work we can learn to accept it.”
In other words, one should just accept that the correlation exists and can be exploited. However, he also mentions specifically that it "seemed to begin to work from about 1997 onward." This actually gives us a major clue.
Correlations between Stocks and "Safe" Interest Rates over Time
Treasury bond yields and stock prices used to exhibit a negative correlation from about 1970 to 1998. This is to say, when bond yields went up, stocks went down, and vice versa. Or putting it differently: the prices of stocks and bonds moved roughly in the same direction.
When the great bull market in stocks finally took off for good in August of 1982, most market participants probably still remembered that the negative correlation between Treasury bond yields and stocks was not in evidence in the era preceding the 1970s. From 1929 to roughly 1969, stock prices and bond yields were positively correlated most of the time – first to the downside, later to the upside.
The positive correlation of Treasury yields and stock prices that pertained in the pre-1970's era returned with a vengeance in 1998 during the Russian/LTCM crisis. It has been with us ever since.
Nowadays it is regarded as perfectly normal that whenever the stock market sells off, Treasury yields will fall and when stock prices rise, Treasury yields will rise along with them (lately this has changed again, probably due to Operation Twist, but by and large it can be assumed that this correlation is still operative).
Next, one must give some thought to how commercial hedgers operate. Generally they will take the opposite position of speculators, i.e. they will short into rising markets and go increasingly long in falling markets. When they are net long the euro-dollar contract, it is because they expect falling interest rates, or rather, because they simply take the other side of the trades put on by speculators who bet increasingly on rising rates as the trend becomes entrenched. Since speculators are mainly trend followers and the commercials mean reversion players, their relative positioning will tend to be at extremes near major turning points.
So what happens these days (at least until recently) when stock prices rise? Usually this will go hand in hand with rising inflation expectations and rising speculative demand for short term credit. Consequently, speculators will tend to increasingly bet on rising rates, with commercials taking the other side of their trades.
Given the positive correlation between inflation expectations/interest rates and stocks, commercial traders are more likely to be net long the contract (a bet on falling rates) near stock market peaks and will tend to decrease their short exposure as rates fall along with a falling stock market.
A 5 year chart of euro dollar futures (weekly candlesticks), with trader commitments. The red line shows the net position of commercial traders. Note the several weeks of discontinuity in 2008 when LIBOR decoupled markedly from the Federal Funds rate and the smaller discontinuity when the euro area crisis flared up in 2010 and 2011.
The Current Situation
By shifting their positioning one year into the future, one is therefore attempting to forecast an underlying rhythm, or time cycle, that influences both stock prices and interest rates and expresses itself as a 12 month lag in the stock market. The euro-dollar contract will by the way most of the time mimic the trend of the Federal Funds rate, but as you can see above, there were brief hick-ups in the 2008 crisis and again during the most stressful moments of the euro area crisis, when LIBOR decoupled from the FF rate. This is because the euro-dollar futures contract also reflects liquidity crises – the spread between euro-dollar interest rates and the t-bill rate, a.k.a. the 'TED spread' is in fact well known as a "crisis indicator."
The big question is of course if this cycle is still operative now that the Fed has adopted "ZIRP." Short term interest rate volatility has decreased because of this policy and short term rates no longer react much to rallies in the stock market. Interestingly, the positioning data in euro dollar futures contract remain highly volatile and these days seem to be far more sensitive to even small movements in the contract's price. So perhaps the indicator will actually continue to work.
A 25 year weekly chart of the euro-dollar futures contract. Here we can see that the contract follows the Federal Funds rate most of the time (just as LIBOR tens to do unless there is a major liquidity crisis). It also shows the pro-cyclical positioning behavior of speculators and the anti-cyclical behavior of commercial hedgers more clearly.
So what does McClellan's forward-shifted overlay chart of commercial positioning in euro-dollar futures and the S&P 500 look like at the moment? Here it is, with a prospective potential path for the S&P 500 indicated as well:
The euro-dollar net commercial position shifted forward by one year and the S&P 500 index. The implied future moves of the S&P 500 are indicated in red – this is of course not a precise forecast, it is merely an indication of what to roughly expect if the correlation discovered by Tom McClellan continues to work.
Now recall our recent post on the "compressed three peaks and a domed house" formation and what it implies about the possible near term path of the stock market (here is the chart). We are struck by the fact that it would coincide almost perfectly with what the correlation between the forward shifted commercial ED futures position and the S&P 500 implies about upcoming stock market moves and turning points.
Not A Forecast
Does this mean one should blindly trust such a forecast? Absolutely not. In fact, it should not even be called a forecast. Similar to what we already said in the post on the "three peaks" formation, such information is not meant to be taken as gospel at all. Rather, what it is potentially useful for is this: if the market begins to behave according to the putative path implied by the "forecast," then one can use it as an additional tool in one's trading arsenal.
The reason why we present such indicators and forecasts here is that they are not followed by a great many people. Very often obscure market indicators prove to be more useful than the ones everyone knows and talks about. Widely discussed indicators are often a bit like with the watched kettle that never seems to want to boil (recall for instance several "Hindenburg Omen" warnings we got over recent years).
Similarly, when thinking about how market fundamentals might evolve in the future and what "trigger events" may present themselves, it is often a good idea to give some thought to potential "left of field" developments that are not widely expected (there exist e.g. a number of "gray swan" situations – this is to say, problems everybody knows about, but that are no longer widely expected to become relevant to financial market action in the near to medium term. Among such "gray swans" is e.g. the potential for a fiscal crisis in Japan, or the potential for war to break out in the Middle East).