Tuesday, April 08, 2008

What of reports of unemployment massing near Sullust?

The reality is we are almost certainly in a recession right now. Friday's employment figures join a long line of coincident indicators supporting that conclusion.

The question is, how much should we really care about coincident indicators at this point? Readers of Accrued Interest are mostly investors, trying to figure out what to buy and sell. We aren't economists at the NBER trying to date the business cycle. Economic releases that tell us about the current state of the economy just aren't very valuable for investors.

Take unemployment. The historic data supports unemployment as a lagging indicator, especially for recoveries. For recessions, it has historically been a coincident to lagging indicator. A quick look at the 2001 recession shows what I mean. The following graph shows unemployment (red) and the S&P 500 (blue) from October 1999 to December 2003.

First let's focus on unemployment. From October 1999 to December 2000, the unemployment rate was between 4.1 and 3.9 every month. No leading indication that the economy (or the market) was about to turn south. During the official recession, March-November 2001, unemployment moved from 4.3 to 5.5.

Even after the economy started to recover, employment kept getting worse. Unemployment didn't actually peak until June 2003 at 6.3%, two full years after the recession was over.

So what's my point? The utility of an economic indicator, as a trader, can come in two ways. Either the number has predictive value, or correctly predicting the number can indicate how to trade the market. In other words, either you can use the number as an input predict future events. Or you can try to predict the number itself and then trade the market accordingly.

The last point is what I call the Crystal Ball test. That is, if you could be given advance knowledge of an economic statistic, say 6-months out, would that knowledge give you a trading advantage?

The 2001 recession shows that unemployment fails this test. Notice the two red shaded areas. In both cases unemployment was stable, but the stock market was falling precipitously. Then in the green shaded area unemployment takes another leg higher, yet the stock market rallied sharply.

Conclusion? During the last recession, unemployment predicted nothing useful to investors. Even had you been given a crystal ball and knew for a fact what future unemployment figures would be, it still wouldn't have consistently indicated the right market trade. In fact it often would have given you the wrong indication.

Of course, there is no sense denying reality. The first Friday of each month, when employment statistics are released, has become a high volatility day. So its a fact that employment is a market mover in the short term. But I strongly caution investors against putting too much weight on rising unemployment. The market is always forward looking, and may already be looking past ugly employment numbers.


Anonymous said...


Employment does not "lead" because it is usually not causal. Recessions are typically caused by excess inventories. If they go on for longer, a contraction in business investment kicks in as a driver. Unemployment is typically the "effect", and it disappears sometime after the inventory/investment cycle turns upwards.

So what's different this time? Obviously, this recession, if we have one, is consumer and not inventory driven. As such, unemployment will switch from effect to cause. Never mind incomes and the need to raise the savings rate; just consider the effect that rising unemployment would have on unsecured consumer credit availability.

Accrued Interest said...

I'd disagree. Isn't this recession caused by excess housing inventories and resulting financial contraction?

Anonymous said...

My point was that excess inventories and declining business investment are not the cause. There is a causal chain leading from housing to credit to real retail spending, and that is what has caused a recession.

Plus, after all, housing is a consumer good.

As far as the credit contraction's impact on businesses, note that even small businesses -- arguably the weakest credits -- are still not complaining about a decline in credit availability (see today's NFIB release). Certainly large corporates are not either. That leaves the "tail" of high-risk corporate credits that has lost access to credit; but its likely this tail represents a much higher percentage of outstanding business credit than it does of business investment.

Anonymous said...

35:1 leverage is the thing that is bedeviling the financial industry.

As in 1929.

Expect breadlines.

Anonymous said...

I'd disagree. Isn't this recession caused by excess housing inventories and resulting financial contraction?

What excess inventory?

Seems most of the "excess" housing on the market was recently lived in.

If the plauge wiped out 20% of US households would you call the resulting recession the result of excess inventory?

Housing was the trigger, but an over-leveraged household balance sheet is the cause. Not enough income to be able to service the debt. The only reason it went on as long as it did was that the housing bubble enabled a debt ponzi scheme. Run up debt on the credit cards and roll it onto the HELOC. Repeat. Refi. Repeat. Roll the HELOC into an option arm.

The recession won't be over until households are able to repair the damage inflicted on the balance sheet over the last decade. Either a quick and painful with bankruptcy or long and drawn out through reduced consumption.

Or maybe the Federal Reserve will destroy the currency and give all debtors a inflation pony.

Either way, the recession won't be over until households have reduced debt and built up a "cushion" to support the next wave of consumer spending.

And that's assuming that the Boomer demographics don't put a long term crimp on consumption in the US.

Christopher Wheeler said...

"Seems most of the "excess" housing on the market was recently lived in."

I read somewhere that 1 out of every 7 houses in the US was unoccupied.

Anonymous said...

I understand that unemployment is usually called the hallmark of recession. And i do agree that unemployment numbers are lagging, at best a confirming indicator of views already expressed by the market. As for the cause of the current recession (which I view has occured/ocurring), it's not sure as important to me as to the effects that it's already causing in the downward negative feedbackloop effects on employment, consumer spending etc etc.

Next thing to do is probably to look beyond the valley.