I am fascinated by the market’s ability to push ahead over the past several weeks while everyday has been lower than its 10 day average volumes. On October 16th we were in the midst of forceful selloff, which looked to be the beginning of a significant move down. However, given that sense, the Fed sent out its minion warriors as it does when the market is not cooperating. And I digress here but I feel this is a useful detour…. Remember that in May 1999 the Fed decided it would be more ‘transparent’ in an effort to give markets more information so they could make better long term decisions. They would do this by providing explicit language about the future stance of policy. What we actually receive is very strategic misinformation in an effort to better control market behaviour.
That proposition is not conspiracy theory but is based on solid economic research. A great deal of research has been done on market expectations and their impact on the effectiveness of economic policies. In a nutshell market expectations often make economic policies (Fed actions) ineffective beyond the short term. And so to combat this dampening effect the Fed implemented a policy of ‘transparency’, specifically of providing forward guidance. Now according to Rational Expectation Model, more transparency about future Fed action would reduce the Fed’s effectiveness and so it would seem odd that the Fed would implement such a strategy. However, providing forward guidance allows the Fed to manipulate market behaviour by forming the market’s expectations for them. Doing so allows the Fed to use the Rational Expectation Model concepts to its advantage.
If you stop to think about the information we receive from the Fed about forward guidance it is intentionally unclear resulting in market confusion over Fed future policy. Specifically, phrases like ‘considerable time’ and whether or not it shows up in the statement has us entirely focused on a completely ambiguous descriptor. Or take for instance the fact that about half the Fed governors publicly discuss their hawkish views while the other half their dovish views. Does this provide clarity to the market as to the Fed’s future action? Of course not. But the misinformation does confuse the market as to what the Fed is really going to do, which Rational Expectation Model suggests allows Fed action to be most effective. Did the Fed have any intention of implementing QE4 immediately after ending QE3 as Bullard publicly described on October 16th? Of course not. However, the misinformation resulted in the exact desired market response. And so we see that the additional ‘transparency’ is really additional misinformation. Ok enough of that, I apologize for digressing, let’s move back to our other topic.
Back to our magical market of low volumes but all time highs. Logically if volumes are high it means the market overall has a view known as conviction. Typically there will be a catalyst for this view and not everyone has the same view but the fact that there is high volume means a lot of people have a view and are thus consequently trading on that view. Now if the market has high volumes and moves up we can say the market traded up with conviction (and vice versa). That is because high volumes suggest conviction and we can see the direction was higher. That’s pretty obvious stuff but what about low volume days? Well low volumes mean there isn’t a lot of conviction in the market one way or the other and so we should see these days move fairly randomly. That’s because the market doesn’t have a view it will tend to trade up or down with equal probability and that holds true for the period between 1950 and 2000.
So with the market on an incredible 6 year bull run one must expect that the market view has been positive, correct? The problem is that when we look at correlations between weekly volumes and weekly returns we get a negative correlation. In fact, since 2009 we’ve had only one year that shows a positive correlation between volume and returns (2012). If we look at the previous bull run from 2002 to 2007 we had only 2 years with positive correlation between volumes and returns. And so we have conundrum in that high volume days have tended down yet the market is at all time highs, which means the market has traded to all time highs on days that should have been randomly up or down. And that seems to defy the concept of randomness.
It begs the question, what force is moving a market to all time highs on low volume? And perhaps we can start by looking at the period since October 16th to date. If we can explain that anomaly perhaps we can extrapolate to the entire past 15 years. Now If you’re hoping I am going to reveal the answer you are going to be somewhat disappointed. I have theories but I really don’t know for certain. What I do know is that this negative correlation between volumes and returns is a phenomenon unique to the new millennium. Have a look at the attached chart depicting volume and return correlations since 1950 (note I used dummy variables of 1 and -1 to depict positive and negative, respectively). We see a shift taking place during the 1990’s and by 2000 we see a clear reversal to negative correlations.