The fixation on second derivative statistics
What? Please stay with me while I do my best to explain the current love affair by most people on Wall Street, aka money managers, mutual fund managers, market analysts, whatever their title and where ever their actual location, with any economic statistic that seems to suggest, “Things are leveling off!”. Once you understand the context and what they are really saying, you’ll have a much better chance to decide for yourself if our economy is really improving.
First Derivative. Any statistic, whether it relates to the economy or any other realm, has several basic properties. The first basic property we apply to any statistic is whether it is up or down from the last data point. Sometimes we use, by convention, last month’s data point if we are looking at an economic statistic. Other times we compare the new data point to the same data point one year ago, again for an economic statistic. These comparisons of economic data, whether compared month-to-month or year-to-year are the first derivative. They tell us the change and/or the percentage change for the statistic compared to last month or last year. Because economic statistics have seasonality, i.e. month-to-month variations related to the time of year independent of overall economic trends, I prefer comparing year-to-year comparisons for economic statistics. Beware of month-to-month economic statistics spread by almost everyone, including the above mentioned Wall Street types as well as the media outlets. Often they state a new data point without telling you what the comparison data point really is, e.g. “Sales are up 1%!” They fail to tell you whether they are talking month-to-month, year-to-year, decade-to decade, or even something else as they wish . Listener beware.
OK, you now know the first derivative for any statistic and specifically for economic statistics.
The Second Derivative. What’s this? Looking at an economic statistic again, the second derivative is simply comparing the current month-to-month or year-to-year percentage change in the data point versus last month’s or last quarter’s or last year’s month-to-month, quarter-to-quarter or year-to-year percentage change. Again, be aware of the what’s being compared. I prefer year-to-year comparisons for all economic statistics, but regardless of the comparison you might prefer, you must make sure you know what’s being compared. If I state, for example, “Today we learned that unemployment is slowing!” What am I really saying? Hard to know based on what was said. What can be known is that this is a second derivative statement, i.e. implying a percentage comparison between the new data point and a previous one, although we have no idea what comparison is really being made. Is this a month-to-month percentage comparison in unemployment change or year-to-year or even something else? Blank. Without a context my assertion about unemployment slowing is meaningless. Hopefully you will be told the context; if not, recognize what you just heard is likely to be just an example of how ’statistics often lie’. Statistics don’t lie, but people often use statistics without context to make their point. Again, listener beware. To make progress here we must all be active listeners, asking the questions left unanswered by whomever is giving us some statistic, especially economic statistics. Hopefully you now understand what I said about unemployment slowing was without a context. Let me modify what I meant to say, ” Unemployment is slowing because today’s data shows 8.8% unemployment, versus 8.5% unemployment last month and 8.0% two months ago.” This, at least, gives you some context for my assertion. This is still a month-to-month comparison, so its not my preferred way of stating economic statistics, but at least its better than no context at all. So now you should know what “Unemployment is slowing!” is trying to say and what questions haven’t been answered by the commentator, regardless of his or her position or title. What is the modified unemployment statement really saying?
The Second Derivative Fixation. The point of this article is to make you understand that most Wall Street people, as defined above, are now using second derivative statistics to make themselves, and hopefully you, more bullish about our economy. Using the above example of hypothetical unemployment data, the data series show unemployment percentages, as they are defined, are currently 8.8% vs 8.5% last month vs 8.0% two months ago. The rate of change month-to-month is SLOWING. What? Well, to be clear, unemployment is still growing (to be exact employment is still declining on a percentage basis compared to previous month’s data points), but the RATE of decline is slowing, at least according to this series of data points. Are things really getting better? No, not based on this statistic, but they seem to suggest that the plunge we have experienced in our economy may not be continuing at the same pace. I recognize this is slightly better than the rate of decline accelerating, but our economy still has not hit bottom. The assumption by people that are fixated on second derivative economic statistics getting ‘better’ is that this new ‘trend’ will continue without pause or reversal and hence our economy will actually be expanding in terms of GDP later this year with employment, i.e. positive job growth, to follow soon thereafter. Wall Street and the media love to push this optimistic view, almost without exception. Are they wrong? Maybe, maybe not this time. The keywords are ‘this time’. Maybe the improvement in second derivative economic statistics will continue and, hence, at some point GDP then employment will get better. I certainly hope and believe things will get better, but I won’t be using an improvement in the rate of decline as telling me its time to bet the farm on the stock market or will lead to, necessarily, a real improvement in our GDP and employment any time soon. Trends last only as long as they last. Even Yogi knows that, even if he stated it as, “It ain’t over till its over!”
One Possible Analogy. OK, here goes. This isn’t pleasant, but, if we bend the laws of physics a bit, we can imagine someone jumping out of a window, accelerating downward toward the street, then, ignoring the laws of physics, suddenly start slowing down en route to the street below. They are still falling, but their speed of descent is actually slowing down. I admit I’m bending the laws of physics here, but the analogy is similar to the current economic statistics. The rate of decline is slowing, but we are still falling. If I carry this analogy to the limit, our jumper will hit bottom, eventually I hope, without getting killed. I hope the same for our economy, but the rate of decline improving (declining?) doesn’t mean we’ve hit bottom yet, nor does it tell us how well we’ll feel once we get there.