Barry Ritholtz reassures us that there’s no recession on the horizon. He could be right, but his argument isn’t. He says that the business cycle follows a predictable pattern: after its drubbing in the latest downturn, the economy slowly picks up steam. Growth accelerates and unemployment falls. Wages rise. Eventually inflation breaches the central bank’s target, and interest rates are hiked to cool things down. That’s when we get another recession. Since wages have barely begun to move, inflation is minimal, and interest rates are scraping bottom, we can be sure that we are a long way from the next slump.
Sounds good, except it imposes a single model on business cycles, when in fact they don’t all fit in one box. There are three different kinds of cycles, as helpfully laid out in an exemplary textbook I’m familiar with. One is the policy cycle, as described by Ritholtz. Yes, that one is flashing a steady green. The second is the investment/profit cycle, whose theoretical basis goes back to Marx, includes Samuelson’s accelerator model, and is driven by the interaction of business costs (including wages), demand, and new investment. The key indicator there is of course profit (and expected profit), and there are no clouds on that horizon at the moment. The third is the financial cycle, of which 2008 was the most recent example. Instability of that sort results from credit growth that props up asset prices rather than increasing revenues or from mismatches between liabilities and revenues. In theory it’s possible to see this kind of trouble in advance, although the actual record is spotty. If we are in for a crunch within the coming year it will probably come from financial forces.
The bottom line is that there are different kinds of business cycles that display different patterns. You can’t show that one type is signaling expansion and conclude that risk is minimal. I don’t put any effort myself into forecasting, and I have no idea how likely a recession is in 2016. People ask me, and I say something like one in three, but that’s just an uninformed prior.