The risk of a meaningful stock pullback’s elevated.

You can tell because… well, have you looked at the historical percentile ranking for the S&P’s current multiple? If not, maybe don’t. It’ll give you vertigo. Especially if you’re already prone to acrophobia.

But if you’re not satisfied with traditional valuation metrics and/or you enjoy indulging in a lot of complicated analysis to “prove” the self-evident, you’ll love Goldman’s revised “framework” for calculating drawdown risk.

As the bank’s Andrea Ferrario and Christian Mueller-Glissmann explained, in a new note, the measure combines “business cycle metrics, sentiment cycle metrics, valuations and leading growth indicators” to derive an estimate of pullback risk “over different horizons.”

The figure gives you some context. Currently, the probability of an equity drawdown of more than 20% over 12 months or 10% over three months is near 30%, which Ferrario noted is “far from previous peaks” but nevertheless “well above its unconditional level.”

If you’re wondering whether this has any predictive value, the answer’s “yes.” Or “yes?” (“I’m Ron Burgundy?”)

Looking back across history, a higher score on Goldman’s probability estimator has generally meant lower forward returns and larger max drawdowns, but it’s far from infallible.

The two figures illustrate the point using different buckets for score ranges. Ferrario and Mueller-Glissmann were quick to note that the signal’s not as strong until drawdown risk moves above 35%, at which point a three-month max drawdown of 10% or a 12-month max drawdown of more than 20% occurred more than half the time.

Coming quickly full circle, Goldman noted that “most of the increase in drawdown probability since September has been driven by business cycle scores, followed by continued equity valuation expansion.”

Higher multiples, the bank reckoned, “tend to increase latent drawdown risk even though they can remain elevated in the near term as long as macro conditions remain supportive.”