The "fear gauge" is wrong

CNBC has called VIX the "fear gauge" for 30 years.

It's wrong.

VIX is ATM 30-day implied vol on the S&P plus a spread. That spread comes from skew steepness and convexity.

When the market sells off, downside skew kicks in. Lower strikes carry higher IV. ATM vol rises. VIX goes up. Even if the entire vol surface is completely unchanged. Not a single strike moved. VIX went higher because the market moved along an existing skew.

Now flip the scenario. If equity skew were inverted (upside skew, like you see in some commodity markets), VIX would rally when stocks went up.

Same index, same math, but a completely different behavior.

The negative correlation between VIX and the S&P comes from the shape of the vol surface. 

This is just pure geometry.

Once you've spent time on a vol desk, it's obvious. Before that, you just repeat what CNBC said.

So when VIX rises 3 points after a 1% sell-off, is that a real signal? Or is the market just sliding along a skew that was already there? You can't answer that from VIX alone.

That's why fixed strike vol is what we actually watch. When implied vol at a specific strike moves, the surface has actually shifted. That's the real read on whether risk is repricing.

See sample from yesterday's briefing in our Alpha Pod: 

VIX just rides the skew.

The market moved. The skew did its job. That's all VIX told you.

Volatility first.