RSSHow to Read Market Data During a Cyclical Turning Point
Construction and macroeconomic data reports are hard to interpret properly at any time but they are potentially more misleading in the neighborhood of a cyclical turning point such as we are now experiencing as the broad economy settles into a slowdown/recession and later, probably late this year, transitions back to another period of accelerating growth.
This is unfortunate because market reports get examined much more closely when the pace of change is close to zero and even the direction of change is not obvious. Why is this? Every data collection process includes some assumptions along with the hard data collected from the market. Typically there is a trending procedure to fill in the blanks when data is missing. Also, most market data are population estimates based on a sample which requires a formula to scale the sample up to the population. These formulas are less accurate when the pace of change in the market is well off its long term trend. For example, at the onset of a recession, the sample of companies stays the same but the population of companies shrinks.
Here are several simple rules for reading data near a turning point.
- Each part of the economy has its own turning point. Data reliability problems are most intense several months before through several months after the turning point. Home construction peaked in February or March 2006. Construction equipment sales peaked in December 2006. Nonresidential construction spending likely peaked about the turn of the year from 2007 to 2008. (I am sure yet; the data is still bad). We expect another turning point for housing later this year when decline turns to renewed expansion.
- When a market enters a turning point date revisions typically become larger and more biased. At the onset of a cyclical low turning point downward revisions predominate because the numbers collectors can not adjust their “fill in the blanks” procedures “instantly. The reverse occurs at a cyclical high turning point. The consequence is that we do not find out about the change in market trends until several months after they happen. You need good sources of real time antidotal information about your market to avoid these surprises. Another good strategy is to anticipate the timing of a cyclical turning point in your market by noting when the turning point occurred in markets that are always earlier in the business cycle. The timing of turning points by market is almost the same in every business cycle except for the few markets that are way off trend such as housing in this downturn and electronics in the last one.
- Long after the fact, when all of the hard data is available and cleaned up, it always turns out that swings in inventory account for most of the deviation of a market or the whole economy from its longer term growth trend. Unfortunately inventory data – as originally reported – is especially unreliable in the neighborhood of cyclical turning points. This is because neither those who report nor those who collect market data pay as much attention to inventory as they do to orders, sales, production or employment. And there is no way for data collectors to audit inventory reports that they receive. For job reports they can check legally required payroll accounting records. But they do not send busloads of people to warehouse with clipboards to count boxes and pallets. So be cautious about inventory data. If you suspect your market is about to hit a cyclical low turning point, assume that whatever amount of inventory is reported is understated – i.e. the inventory problem is worse than thought and will have a bigger negative impact on sales than the official inventory count implies. Of course this process also works in reverse at the onset of a market recovery. For homebuilders and their suppliers this means that the eventual turn to recovery will be more abrupt and pronounced than is generally expected in light of the current huge overhang of surplus homes for sale.
