Friday, March 2, 2018

Questions America Wants Answered "Was There A Stock Market Bubble? Did It Burst?"

From The Capital Spectator:
It took longer than usual, but the priced-for-perfection bull run has taken a body blow in recent weeks. That’s good news for investors with relatively long time horizons since lower prices equate with higher expected returns. Meantime, with the benefit of hindsight, let’s consider how a pair of econometric tools have fared for monitoring bubble risk and bull/bear market regimes. The question before the house: Have these metrics provided any value for managing risk in an effort to cut through the noise and anticipate trouble based solely on the hard data? As a preview, I’ll argue “yes,” but with caveats.

Several years ago I reviewed the potential for using a rolling 36-month ADF test to quantify the state of bubble risk for the US stock market (here’s the 2014 article). Soon after I analyzed an econometric measure for objectively estimating if equities were in a bear or bull regime via a statistical framework known as a Hidden Markov model (HMM). Although the historical results on both fronts looked encouraging for anticipating critical turning points in the market, the out-of-sample data would ultimately decide if these tools were useful in the real world.

In the years since I published those two pieces I’ve periodically looked at how the models have stood the test of time. Last November, for example, I noted that the elevated readings for bubble risk appeared to suffer false warnings. Despite the warning, the S&P was still rising, and with an unusually low level of volatility.

Several months later, however, it’s reasonable to wonder if the recent round of high bubble risk readings were accurate after all? The warnings were early, but one could argue that the probability was high that a sharp correction was lurking. The only question: What would be the catalyst and when? Recent events appear to have solved the mystery....MORE