There was no substantial edge apparent throughout the 70s, but starting in 1983 through 2009 there was a substantial bullish tendency. The last 16 years this week has seen more of a struggle. But there remains one day during Memorial Day week where the upside tendency seems to have persisted. That day is Thursday.

A 70% win rate and a strong, steady move higher suggest Thursday looks like a pretty good day from a seasonal standpoint. Traders may want to keep this in mind for later this week.

Related Quantifiable Edges Studies

• An updated look at Thanksgiving Week Stats
• The Reversal Tendency of Labor Day Week
• Will Tomorrow Be Another Bullish Holy Thursday?
• Twas 3 Nights Before Christmas (Russell 2000 version)

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

How about a free trial to the Quantifiable Edges Gold subscription?

Shaded rows are those instances where there was less than a 20% drawdown at the start of the rally. The current instance saw one of the shallower drawdowns. Only 1955, 1976, and 1982 saw rallies that were not coming out of larger drawdowns than the current one. After such strong rallies whether they were bear or non-bear, the next month generally did not see much progress. But longer-term results were quite impressive. After the 19 prior 6-week rallies, SPX averaged +18.8% over the next year with a 94.4% win rate, versus a +9.4% / 74.5% baseline. Below you can see the signal versus the baseline numbers.

Signal vs baseline forward returns (top 20 rallies)

Lastly a little graphic showing the one-year forward returns after each of the 6-week rallies:

So the momentum over the recent six-week period seems to be positive. We’ve seen this other ways but I thought this was an interesting look. Also impressive is that looking at the eight non-bear instances the average drawdown was only 8.29%, while the average run-up was 18.96%. There are certainly risks out there, but strong moves like we have seen recently tend to be a positive over the intermediate and long term.

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

How about a free trial to the Quantifiable Edges Gold subscription?

NAAIM has published its “NAAIM Exposure Index” since 2006. I did some research a few years ago on the index to determine whether the numbers might be valuable as part of a model. I found that 1) strongly oversold readings could be indicative that the market is so oversold it is ready to rally, 2) strongly overbought readings are not a contrarian indicator. In fact they often suggest strong momentum that is likely to continue. I had a discussion with some NAAIM members a few weeks back. The topic was the Exposure Index: who used it, what they used it for, and where was the actual value in it?

So I decided to revisit my old research and see if I could create a model based on those concepts that could demonstrate the value that the index provides. Results came out better than anticipated. Below is a look at the profit curve of the model versus the S&P 500.

The NAAIM Exposure Index is released each week during market hours on Thursday. So the model trades at the close every Thursday afternoon. (Thanksgiving and other holidays will sometimes bump it forward one day.) There are three setups that will cause the model to go long.

This week’s reading is among the bottom 10% of readings seen over the last 52 weeks (one year). Additionally, it is not lower than last week’s reading. Essentially this is looking for a strongly oversold reading, but not looking to catch a falling knife. So we want to see the Exposure Index near or slightly above where it was the week before.

This week’s reading is at least 80 and it has risen at least 15 points over the last four weeks. This is looking for strong and growing enthusiasm.

This week’s reading is at least 100. As it turns out, when you have a bunch of smart investment managers getting leveraged, there is a good chance that they are right and that there are market gains ahead.

If any one of the above is true, the model will go (or remain) long 150% SPX at the close on Thursday, and will remain long at least until the following Thursday when the new number is published. Below is a stats table summarizing the results.

The numbers here compare very favorably to the SPX. Leverage when there is an edge and cash when there is not. As you can see the model’s only invested 19% of the time. That is a big risk reducer. One thing to note is the Max Drawdown. It is shown as less than 10% in the table above. But that is partially because this model is just using weekly closing numbers. I created a version in Realtest that uses daily pricing. And thanks to March 2020, the max drawdown reached 25% using daily bars. (Still less than half that of the S&P 500.)

Is this a model I would actually trade? No. I don’t think I’d trade any model where I had to hold for a full week based solely on survey results. Of course the fact that it is built solely on survey results, and does not take into account price, or trend, or volume, or breadth, or anything else, is what really demonstrates the value of the NAAIM Exposure Index. And that was kind of the whole point. Trading solely on the index may be overenthusiastic but utilizing it as one input within a larger model seems completely reasonable and could very possibly strengthen it.

I have the model available in both Excel and Realtest format. If you have a username and password at Quantifiable Edges, then you can download the model from the Other Code and Spreadsheets page. If you don’t have a log-in, you can get one by simply signing up for a free trial! And if you don’t want a free trial right now, then just join our email list and you will receive a copy. Easy as pie.

Disclaimer: The performance shown above is hypothetical and does not represent an actual trading account. Results were generated frictionlessly, meaning they do not include commissions, slippage, or other costs that would be incurred in live trading. Past performance, actual or hypothetical, is no guarantee of future results.

Of the 25 instances that rose on the first day in May since 1987, 17 of them closed lower 4 days later. Below is an equity curve that shows how it has played out over time.

I’ll note the big drop in instance 17 was the 2010 Flash Crash. Even without that instance there appears to be a solid downside inclination.

I will also note that while seasonality appears negative over the next few days, there was quite a mix of bullish and bearish studies featured in the subscriber letter over the weekend. I don’t view this as a reason to trade on its own, but it is a tendency that traders may want to consider.

Is it still potentially bullish? I believe so. Is it the layup that it once appeared to be? No. But I still find it interesting and perhaps worth consideration when thinking about Wednesday’s potential.

The time around Thanksgiving has shown some strong tendencies – both bullish and bearish. I have discussed them a number of times over the years. In the updated table below I show SPX performance results based on the day of the week around Thanksgiving. The bottom row is the Monday after Thanksgiving week. The top row is the Monday before Thanksgiving.

Monday and Tuesday of Thanksgiving week do not show a strong, consistent edge. But the data for both Wednesday and Friday looks quite strong. Wednesday has had the most consistent gains (76.9%), and the largest average p/l of 0.3% ($297.44 on $100k.) Friday’s stats look fairly impressive, but it has not performed nearly as well over the last 15 years as it had previously. Meanwhile, the Monday after Thanksgiving has given back a good chunk of the gains that were realized on Wednesday and Friday. The worst “Monday After” came in 2008 with a 9% 1-day decline. Since then, Monday After performance has been basically breakeven.

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

How about a free trial to the Quantifiable Edges Gold subscription?

The 4-day numbers are basically inverted. So the Tues-Fri after Labor Day have not seen any consistency without the delineator. But the trend filter reveals a striking difference. Over the last several weeks the market has rallied nicely, so we are currently facing the 2nd scenario. Below is a look at the profit curve for the 4-day exit following the current setup.

Choppy but an obvious downslope. Traders may want to consider this for the upcoming trading week. Have a great Labor Day!

To see other studies featured in this weekend’s subscriber letter, check out a free trial.

As you can see it is the only month with a negative Day 1 return. Below is a look at how it has played out over time.

This doesn’t appear to be a bearish setup. Just a bunch of chop that has failed to generate gains over time. But while it doesn’t hint at a down day on August 1^st, it certainly shows that August 1 has certainly not been seasonally bullish.

Interestingly, 36 of the 49 instances (73.5%) actually saw the SPX realize gains while the Death Cross was in effect.  The problem is the losing trades were very large.  And even most of the winners saw a sizable round-trip lower before they were able to carve out some gains (like the last one in 2022).  The average drawdown for these 49 trades would have been 13.2%.  And there were 5 separate instances that saw drawdowns of at least 45%.

Even though the giant losers were relatively rare, their impact is large.  And the fact that the 1^st instance was the worst instance also provides a great example of how devastating large drawdowns can be.  The profit curve below shows a hypothetical portfolio of only being invested in the market during SPX Death Crosses.

The 1^st instance from 1929 – 1933 saw the portfolio rise to $120k before falling down as low as about $20k and then finishing that trade with a value of about $34k.  And it has never managed to get back to breakeven.  The 73.5% “win rate” on the Death Cross tells me it is NOT a reliable timing device.  But the few instances of massive losses show just how valuable it can be to protect gains and avoid large portions of nasty bear markets.  If we get into a big bear market, I won’t actually be sitting out of the market for several years.  But it does allow us to adjust strategies, exposure and other risk parameters.  The Death Cross / Golden Cross on its own is not a great system.  But it can help us put the market into a context where we can better evaluate opportunities.  And I have also found it helpful when combining with other timing indicators, as I’ve done in the Market Timing Course.

Note: The Quantifiable Edges Market Timing Course does look at the Golden Cross / Death Cross in combination with other timing indicators.

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

Those are some impressive stats over the next few days. I also produced the 3-day profit curve.

That is a persistent move from lower left to upper right, serving as some confirmation of the upside edge suggested by the stats table.

Want research like this delivered directly to your inbox on a timely basis? Sign up for the Quantifiable Edges Email List.

How about a free trial to the Quantifiable Edges Gold subscription?