Yesterday, the ISM Manufacturing index, one of the key US economic indicators, dropped a massive five points in January. This was a major upset since economists expected only a moderate drop of one point, according to Bloomberg survey data. The surprise fall, of five times the expected magnitude, caught investors off guard. All major indices (Dow Jones, DIA, S&P, SPY,) retreated more than 2%, while the VIX-index spiked. But just how bad is this drop in the ISM index in the longer term?
To be clear, in deciding the longer term impact of an ISM shock, I will pay no attention to the usefulness of the ISM index as a key leading indicator, or to the methodology used to calculate it. Since the ISM was published yesterday there have been a lot of news flashes that doubt whether the ISM index is able to capture the developments in all US businesses. Especially since only 400 member companies take part in the monthly ISM survey. Furthermore, I will stay clear of any anecdotal evidence used to explain the big drop in January. This because there is always some story to be told about why the index number deviated so much from the expectations. Besides big deviations can also have something to do with the quality of the expectations. Finally, while I do realize that the size of the surprise is probably, in some way, linked to the reaction on the markets, I do not take this into account explicitly. That said, a drop of five points is rare and probably comes as a surprise in almost any case. I just want to find out if a drop of this magnitude impacts the outlook for equities. To determine this I will look at history.
I just mentioned that a fall in the ISM Manufacturing index is rare. To make this more tangible I have calculated the historical probability of a drop of more than five points since its inception in 1948. The results are shown in the table below. Only in just over 3% (or 24) of all months since 1948 has the index fallen by more than five points. A drop of this magnitude certainly does not happen every day.
Now quite some people, many of them economist, see the unexpected negative result in January as a strong indication that the ISM will fall further this month. Therefore I added the historical probability for months in which the ISM index fall five points or more and are then followed by another month in which the index goes down. Out of the 24 months in which the ISM did lose five points or more, 14 (58%) are followed by another negative month. So, while true that the probability of another drop is more than 50%, is it by no means extreme.
What does it mean for equities? Well if anything a large drop in the ISM is positive for equities. The graph below shows the average return on the S&P 500 index 6 months after the ISM was published. For all months the average 6-month return on equities has been 4.4% since 1948. The second column shows the average 6-month return on the S&P 500 for the 24 months in which the ISM index dropped more than five points. As can be derived from the graph, that average of 5.9% is actually higher than the overall 6-month average. Investing every 6-month period after a five point drop in the ISM would have earned you 1.5% additional return. But that's not all. If the ISM index does indeed drop another month, the average 6-month return gets even better. Investing only in the 6-months after the ISM index falls by five points or more, and the next month turns out negative again, in the index would have yielded an average return of 8.4%. That is a full four percentage points more than the overall average of all 6-month periods.
I realize this is a very simplistic way of looking at the impact of big drops in the ISM Manufacturing index. The details of why this happens are different every time. Also, a sample of 24 is not all that much which makes the data a little shaky. But then again the data does reveal some interesting results. Independent of what factors made the ISM fall, equities have on average done pretty well afterwards. And, while probability is relatively high that the ISM index will fall again this month, this has not been bad for equities. Historically that is.