Skip to main content

Party Like It's 1929

Correlations between 2020 and prior post-crash environments remain sky-high. Here's an updated look at the data.

2020 continues to trade in lock-step with a couple pretty ominous years.

We’ve been following these incredibly high correlations between 2020’s market environment and the post-crash markets of 1929 and 1987 for a while now. And the takeaway probably isn't what you think.

(Click here for the last time we took a look.)

Let’s just cut to the chase with the latest chart:


In the chart above, the dark blue line shows 2020’s price performance, with 1987 in blue and 1929 in red.

(More on the chart at the end.)

A couple of important takeaways – first, while the magnitudes of the moves are slightly different, the turning points of the major trends clearly synch up incredibly closely.

And second, we’re at the point now where those two prior crash years decoupled from one another and led to very different longer-term outcomes.

The chart above also raises a couple of questions – like why is 2020 trading like these prior periods in the first place?

I think it’s important to note that I don’t think that the chart above suggests that we’re due for a 1929-like market going forward.

Having sky high correlations with totally unrelated prior periods doesn’t somehow lock us onto a track pointed at the exact same outcome stocks experienced back then.

That would be nuts.

Instead, the thing that these periods have in common is a lack of idiosyncratic information specific to the crash. In other words, we’re still experiencing a behavioral reaction to a colossal selloff from February, not a reflection of the current economic realities (whatever they may be).

Investors are still on autopilot right now.

And when 2020’s market does reprice, it could come with a jolt.

Like I said a moment ago, we’re right at the point now where those prior years decoupled from one another. That’s worth keeping in mind right now as you weigh your allocations heading into a test of new-high territory for the S&P undefined.

There’s a whole lot more uncertainty in this market than the price tag would suggest.

A couple of important/interesting post-scripts:

1. The Fed

The timeframes above aren’t totally identical. The Fed’s aggressive stimulus campaign has sped up the post-crash rebound in 2020 to a pace we’ve never seen before. And they’ve signaled that it’s not going anywhere anytime soon – which is another good reason not to expect 2020 to continue to track 1929 or 1987 perfectly.

2. Chart Crimes & Correlations

I’ve seen some other charts on social media recently trying to show analogs between 2020 and 1929 that use much longer timeframes. The ones I’ve seen are problematic because they use different price scales, which takes away most of the meaning.

Time, on the other hand, makes sense to adjust when looking at historical analogs. In the chart above, we scale the 1987 and 1929 time axis enough to synch up the initial price drawdown and account for the Fed’s added impact on 2020. Why do that? 

In plain English, investors don’t act differently because “we’re 34 days from all time highs”; but they do act differently because “we’re 34% from all time highs”.

Finally, for a little extra quantification of the relationship we’re seeing to 1929 specifically, here’s the current rolling correlation between now and then:


While the relationship has been fading a little bit from its peak back in March (during the market’s lows), it’s still incredibly elevated, clocking in at 0.81 at last count.

I’ll keep an eye on this relationship in the weeks ahead. Watch this space!