Recession Dating: What the Investor Needs to Know
Today the private National Bureau of Economic Research (NBER) revealed that it was unwilling to call the end of the recession.
Does this have any relevance for investors? No.
Why the NBER Seems Out of Touch
There is a clear distinction between what we all see in the world around us and the NBER dating system. Let us make this explicit.
Most of us go by "feel" augmented with a few selected data points. There is a trend line for economic potential. It implies relatively low unemployment, solid GDP growth, good corporate earnings, and (sometimes) budget surpluses for government. The current recession started to "feel" bad before the NBER declared the start. It still feels bad even though things have bounced a bit. We all know that the employment and housing situations are terrible and that the economy is far from a real recovery.
One popular conclusion is that the leading economists on the committee are a bunch of clueless bozos. This approach was especially popular a few years ago when CNBC was asking anyone and everyone if they thought that "we were in a recession."
Let's make this as simple as possible. The NBER is attempting to identify peaks and troughs in the economic cycle for research purposes. To emphasize:
- It is designed to facilitate economic research;
- The starts and endings of recessions are peaks and troughs in the economic cycle;
- None of the analysis is related to a comparison with the economic potential.
Briefly put, the NBER has a research mission, not a policy mission. Mark Thoma does a nice job in making this distinction and raising the question of whether the NBER should attempt more timely calls.
The NBER method is important because everyone uses it to analyze past recessions. Even those who disagreed with the NBER on the timing for the start of the recession now use that date in their analysis. It is worth a closer look.
In April of 2008 I wrote an article explaining the process and suggesting that you could probably win a contest about the start of the recession by predicting October or November of 2007. I missed by a month. Check out my handicapping of the entrants.
The key to the NBER method is that they need to see a really significant move in several indicators. While it is an over-simplification, I like to think of this process as involving a "confirming event."
The 2001 recession was eventually dated as starting in March. The confirming event was 9/11. If 9/11 had not occurred, there might have been an economic bounce that would have "canceled" the March peak and established a different peak. Here is what the NBER said:
Q: The NBER has dated the beginning of the recession in March
2001. Does this mean that the attacks of September 11 did not have a
role in causing the recession?
A. No. Before the attacks, it is possible that the decline in
the economy would have been too mild to qualify as a recession. The
attacks clearly deepened the contraction and may have been an important
factor in turning the episode into a recession.
I would like to emphasize a metric that I believe to be novel: The time between the economic peak and the "confirming event."
Think about this for a moment. If the attacks had come sooner (or later) the timing of the recession confirmation would have been different. This is important because the confirmation date is generally the start of the government reaction.
This has direct implications for the current recession. The key date is not the official NBER date. It is the date of the confirming event. The collapse of Lehman created the complete cessation of normal lending, the firing of workers, the canceling of expansion plans, and the end of mortgage securitization.
Thinking about Timing
While it is difficult to understand — we are all blinded by hindsight — the Lehman failure changed everything. I realize that it is very difficult to grasp a counter-factual world, one where the crisis was handled better.
Instead let me suggest this minor challenge. If Bear had failed instead of Lehman, the recession would have been confirmed in March instead of September. Policy actions would have started sooner and the trough would be sooner. The recession would be shorter.
If Lehman had been "saved" the confirming event would have been delayed. For how long I do not know, but the Lehman timing could not have been worse. It happened at the end of a lame-duck administration. Even when legislation passed, the key decisions were on hold for months during a Presidential transition. There was further delay as the Obama team fumbled the ball on key appointments and policy decisions.
My point? The change a few weeks in timing — either way — could have made a big difference.
The biggest lesson for investors is to ignore all of the charts about how this recession compares to those of the past. There are not enough examples for inference. More importantly, the timing before the confirming event has too much variation.
My own approach is to start the shot clock from the time of the confirming event, the fall of Lehman. Here is a nice chart from Calculated Risk. It shows the recessions centered on the trough.
You will have an investing edge if you think of the recovery in terms of time from the bottom, rather than time from the start.