Stock markets that do well in the first half of a year are significantly more likely to do well in the second half. This should not surprise us, because these things do move in trends.
We often look to past data to predict future trends, although how useful this ends up being will depend in large part on the quality of the past data we use. This can in certain circumstances add to the level of confidence we have going forward, as long as we understand that today’s markets and yesterday’s markets can differ quite a bit, and therefore we need to be aware of the limitations involved and ensure we remain sensible in interpreting these comparisons.
There are many examples of people trying to apply past data to the future in ways that are suspect, and the recent concern that low yields on bonds is a good example of this. Just having a raw correlation isn’t enough, and we actually have to be pretty careful with correlations between two distinct events due to the possibility of other conditions driving the change instead.
This is such a great example of how things can go wrong actually, because in this case the other condition is actually what is sought to be predicted by this correlation. When we deteriorate into a recessionary period, this economic climate causes both bond yields and stock prices to go down. In these cases, we will see both happen, and generally the bonds are the first to go since they are so sensitive to economic conditions, with stocks following later as the situation deteriorates and this ends up affecting the forces that drive stock prices in a negative way.
It’s not hard to figure how this happens, and with the bond yields falling, we can bank on this because moving toward a recession means inflation drops and yields always drop in turn, because they track inflation directly.
With stocks, such economic declines influence stock prices as well, and although some of these effects are direct, such as people having less money to invest and needing to cash out their investments more due to things such as losing their job, most of the effect here is through changing investor sentiment and the negative momentum that this causes.
We then see bond yields drop, and wonder if this will cause stocks to drop as well and whether this means a recession is coming. It might be the case that the two go together more often than not, 3 out of 4 times for instance, but that is not enough information to come to any real conclusion. In these 3 out of 4 cases, we have a recession coming and it comes, and in the fourth, this is independent of a recession.
In a given case, like the one we’re in now, this might seem to indicate that there is a 75% chance of a recession, but the truth is that we have no idea of what may be in store just based upon this data because we have not distinguished between the two instances. This is not at all like drawing a marble out of a bag that has 3 red marbles and 1 green one and assessing the probability of a red one based upon that.
If We Are Looking for Causal Relationships, We Need to Think Causally
Real life situations are more complicated than this, and even though we may know that a recession comes more often when we get these low bond yields, what marbles are in the bag differ depending on the circumstances, and in this case, depending on what is causing the yields to go down. Instead of just being satisfied with this correlation, we need to look into the bag to see what marbles are actually there this time, which means looking at both what is causing the yields to drop as well as the outlook for a recession independent of these yields.
Using this correlation actually gets this backwards, as if it is the coming recession that causes both, it is the coming recession that needs to be confirmed, not the presence of a variable that it causes but one that may be caused by other factors as well.
Whenever we look to analyze things, we cannot just be satisfied with a particular conclusion if we can further distinguish our analysis, and this certainly applies to analyzing future market behavior. In this instance, we know that if we move into a recessionary period, yields will drop, but yields dropping does not portend a recession, and there are many examples of this.
We see the yields drop and then we ask ourselves why, and this takes us back to assessing the likelihood of a recession independent of these yields. The lower yields therefore add nothing to the discussion.
Even if they did, and even if we knew that stock prices will be dropping, and a recession is coming, this still doesn’t mean that it is time to sell our stocks. Since our goal is at least supposed to be to look to stay in our stock positions during bullish times, if we’re still in a bullish period, this would be a rather foolish time to get out, and even those who are prone to exit quickly need to ensure that they are doing so when it at least makes sense to do so.
We see a lot of this talk of the market being tired these days, as well as a bear market coming sooner or later, but if we don’t know the timing of it, whether it be sooner or later, if it is not now, and things still look pretty good, it is also foolish to be too hasty. The people who got out back in March or in May on the dips that we saw are probably not too happy now unless they got right back in when the move showed its hand.
Given that we did put in a very good first half of 2019, this should in itself bode pretty well, but we also need to be careful about how much weight we put on these things. 17% in a half a year is pretty impressive though, given the average market return per year is about half of that, so this is 4 times the norm and the best first half since 1997.
Whenever we see a positive first half of any magnitude, the so-called odds of the second half being positive as well goes up, and we could use this data to say that it is 60% more likely to have a positive second half when this happens. When the move up has been this substantial, this correlates even better with a positive second half.
The question we need to be asking whenever we have correlations like this is whether or not it makes sense to say that the bullish first half caused a bullish second half in some way, not by way of a direct cause of course but by being directly influential. In this case, the answer is that this does make sense, because of the nature of markets.
This Data Makes Sense Because There is a Causal Relationship to Some Degree
Markets are driven primarily by momentum, and therefore when we have it, it’s more likely to persist. This doesn’t mean that it will, and all trends do come to an end, but it’s more likely to persist than not, and this is actually the underlying basis behind all of investing and trading that is done with a positive expectation.
This does not absolve us from looking ahead and seeking to distinguish between this instance and other ones, and this correlation really only serves to quiet the talk about the market looking weak as some hold. Ultimately, as long as things keep moving up, staying with the move makes perfect sense, but we also want to be careful about casting our vision too far down the road.
As we travel down it toward this view, in this case the end of 2019, conditions can change which should alter our view. If there is only a 40% chance of rain for the day and you’re going to be outside, and you don’t want to get wet, if the rain does come, you shouldn’t be standing out in it unprepared and cursing the weatherman. The right way to use this data is to expect that it probably won’t rain but also be prepared if it does.
We do need some sort of outlook though if we are investing, although predictions of where we may be in 6 months aren’t really as helpful as we might think, and actually not all that helpful at all. We do need a plan that will allow us to decide these things as they go, but deciding 6 months ahead in this case doesn’t really make sense.
While we should never be looking to trade on any forecasts of this length, this can serve to put us more on alert if we are the sort that needs such things, even though investors should always be on alert anyway. What we are looking to distinguish here is the possibility of trends reversing, the bear market in this case coming to pass to an extent where we should react to it.
It can be fun to guess at where we may be at the end of this year, or in a few years, and these can be fairly well-educated guesses, but we are still guessing. If we are guessing things will remain the same, then staying the course can be made easier, and if we guess that things may turn around, this can also make it easier to get out if the time does come, but in both these cases we still need to wait for the situation to unfold enough before it makes sense to act.
In regard to our current correlation, we can sit back and be comforted somewhat by the fact that we’ll make more money in stocks in the second half 6 out of 10 times, and those odds look decent enough, but this should never have us taking our eyes off the road and to ignore situations that may distinguish this current instance from others.
Right now, there’s no real reason to make such a distinction, in spite of the concerns that some people have about the economy, trade, earnings, the Fed, consumer confidence, bond yields, or what have you. From where we sit right now in early July, the rest of the year looks pretty rosy, but that’s right now, and this is certainly subject to revision.
If it does need to be revised, what has happened in past years during a strong first half will not matter, and this 60% distinction doesn’t matter right now other than this having a positive effect on investor confidence. Investor confidence is what this is all about, and if we just watch that, we’ll be pretty well informed without the need to look to the past at all. Investor confidence can cause more investor confidence, and this is the reason for this positive correlation, but only if it is maintained.
Investor confidence is expressed in price trends, and we’re in a pretty good one this year, so far anyway. We can make these things way more complicated than they need to be, but it really does come down to where the market is headed and this is what makes it rain when it does, when we lose that.
Therefore, the skies look pretty clear right now based upon that, regardless of what may have been going on during those other years. A sunny first half does often mean a sunny second half as well, but only if the sun continues to shine.