The Next 3 Weeks Could Determine How Stocks Perform This Summer

Summary

• From a M2 monetary standpoint, weeks 20-22 of each year have the power to set the stage for how stocks perform through the summer.
• A best-case scenario would see M2 climb for weeks 20 and 21 and retreat less than 1.2% for week 22.
• A bull market would probably continue in that case into 2020.
• Worst-case scenario, M2 falls for weeks 20 and 21, continues to fall more than 1.5% for week 22. If that happens, I plan to go short by mid-May.
• Most likely, something in the middle and more indeterminate will happen, but we'll soon find out.

I've written a lot about how steep changes in the money supply growth rate can predict, and, in my opinion, cause precipitous declines in the stock market. This time I'm going to try to get very specific in what to look for in order to determine what happens past July this year when conditions typically get dangerous for equities. I'll spoil the ending first: The next three weeks of M2 money supply data could determine whether we see smooth sailing for stocks this summer or whether we're going to see a continuation of the bear market we saw late last year or worse. Now, I'll try to explain why and how with hard numbers and how I arrived at them.

A short background: Three serious market declines since 2008 have coincided with either very slow or even absolute shrinkage in the M2 money supply over a quarter. In 2008, the quarterly M2 average money supply shrunk absolutely throughout all of September (data here at table 2). That hasn't happened since. The crash occurred towards the end of that month. The flash crash of August 2015 coincided with the lowest M2 growth since 2008, at just an annual rate of 1% by the end of July. The average is usually around 5% to 6%.

The most recent decline late last year was under slightly different circumstances. M2 growth did not dip below 2% at its lowest last year, but it also had not grown at an annual rate higher than 6% since week 22 of 2017. By comparison, in all of the 5 years previous, M2 growth was above 6% for most of the year. Combined with the massive rally of 2017, there was not enough liquidity added in 2017 and 2018 to sustain new highs and so we saw a sharp decline at the end of 2018.

The timing of the most recent decline was impossible to predict using monetary data alone because the fall did not coincide with a sudden stoppage in M2 growth as did the late summers of 2008 and 2015. It was more of a cumulative effect that eventually bled through. However, this year, we may be headed for another full stop in M2 growth and perhaps even absolute shrinkage by late July when seasonally M2 growth rates hit their seasonal lows each year. Then again, we may not. It all depends I believe in the next 3 to 4 weeks of M2 data.

What I have done is extrapolate M2 growth rate possibilities out to the third week of September by applying the week-to-week growth rates of the past four years to this current year. Speaking first in broad terms, what I found is that if we see similar week-to-week changes as we did in 2016 from now until September, we should be able to cruise through the summer without much damage, and we could even see sustained new highs. If we see the same pattern as we did in 2016, M2 growth should bottom out at 2.6% annual, which is not that bad.

2017 rates come in second place, with the rate bottoming out at 1.4% by the end of July in that case. 1.4% is a dangerously low level, but recovery is possible. Applying 2018 week-to-week growth rates from here would see a low of 1.2%, similar to what we saw in 2015 when growth bottomed out at 1%. There would likely be some sort of flash crash in that scenario by August but probably nothing major for the economy as a whole.

However, if we apply 2015 growth rates from this week of the year until mid-September, stocks are in for serious trouble, and we would probably be tipped into a hard recession and some kind of crisis similar to 2008 by the fall. Applying 2015 week-to-week growth rates from where we are now would have M2 growth stop completely by late June and then go into reverse for 4 weeks in a row. From a monetary standpoint, this would be worse than 2008 when M2 shrunk absolutely for only 3 weeks in a row. The chances of a serious decline in stocks and the broader economy, in that case, would be very high. I for one would be eager to go all in short the major indexes (DIA) (QQQ) (SPY) if that happens.

But how can we know the M2 trajectory before it happens? The answer is you can't with certainty, but for the reasons below, the next 3 weeks could determine where we'll be by July and August.

The table below shows every weekly M2 number for each year going back to 2015 from this week of the year until mid-September. These are 1-week averages, not quarterly averages, so they oscillate up and down, which is normal. It is the 1-week averages that determine the quarterly averages over time, and therefore, the quarterly growth rates. You can find the raw data for all this in the extreme right column of table 2 at the Federal Reserve's money stock measures report each week for weeks 19-34 of each year. Interspersed between the raw M2 data is the percentage change week-to-week (W2W).

Week #	2018 M2	% Change W2W	2017 M2	% Change W2W	2016 M2	% Change W2W	2015 M2	% Change W2W
W19	14126.7		13,539.3		12,777.8		12,082.0
W20	14122.9	-0.03	13,588.9	0.37	12,791.8	0.11	12,061.6	-0.17
W21	14146.3	0.17	13,626.8	0.28	12,801.7	0.08	12,005.2	-0.47
W22	13989.5	-1.11	13,424.9	-1.48	12,649.7	-1.19	11,824.6	-1.50
W23	13898.9	-0.65	13,406.8	-0.13	12,663.8	0.11	11,877.3	0.45
W24	13961.4	0.45	13,430.7	0.18	12,685.7	0.17	11,885.6	0.07
W25	13979.0	0.13	13,449.9	0.14	12,703.4	0.14	11,900.6	0.13
W26	13968.5	-0.08	13,427.5	-0.17	12,652.2	-0.40	11,854.9	-0.38
W27	13953.0	-0.11	13,408.1	-0.14	12,646.1	-0.05	11,915.7	0.51
W28	14072.7	0.86	13,505.4	0.73	12,765.2	0.94	11,953.3	0.32
W29	14083.0	0.07	13,498.0	-0.05	12,799.9	0.27	11,969.5	0.14
W30	14099.2	0.12	13,477.8	-0.15	12,771.1	-0.23	11,915.4	-0.45
W31	13983.5	-0.82	13,403.0	-0.55	12,702.5	-0.54	11,934.8	0.16
W32	14100.3	0.84	13,521.8	0.89	12,839.5	1.08	12,068.4	1.12
W33	14138.1	0.27	13,581.9	0.44	12,839.5	0.00	12,026.2	-0.35
W34	14131.5	-0.05	13,573.8	-0.06	12,844.6	0.04	11,992.2	-0.28

Click to enlarge

In weeks 20-24, 28, 30, 33, and 34, week-to-week growth (or shrinkage) rates diverge between years, with the heaviest resulting impacts on the differences in quarterly growth rates down the line in the earliest weeks. Still, despite divergence, you'll see that seasonal patterns persist. For instance, week 22 always sees a sharp reversal and week 32 strong growth. This helps assure that the data is lined up correctly year-to-year and serves as a reminder that we won't see wild divergences where, say, one year we see 1.5% growth and another 1.5% shrinkage for the same week. Seasonal patterns repeat, just not perfectly.

Since later weeks build on the total for previous weeks, the earlier the week, the greater the impact going forward. We have just finished Week 19, and as you can see above, the next five weeks are all fairly divergent between years and will have the greatest impact on quarterly growth rates going forward, starting with this week, data out at 4:30 PM if you're reading this on Thursday, April 18, 2019.

As noted above, 2016 growth rates would be the best-case scenario, followed by 2017, 2018, then 2015 as a worst-case scenario. And as you can see, in 2016, we had positive growth rates for 4 out of these 5 upcoming weeks, and a relatively shallow retreat of 1.19% in week 22 compared to other years. In 2015, we had, by contrast, shrinking growth rates for weeks 20 and 21 and a sharp retreat of 1.5% in week 22. Those three weeks sealed the deal for 2015, as recovery in week 23 already wasn't enough to save the year. 2017 had a deep retreat in week 22 of almost 1.5%, but that came from a strong base built off of weeks 20 and 21 where M2 was still growing strongly compared to other years, so 2017 turned out OK. 2018 had the shallowest of week 22 retreats at only 1.1%, but then had a strong decline in week 23, unlike the other three years, making it the second worst of the last 4 years.

The patterns here across the years lead me to believe that the next 3 weeks in particular, and to some extent the fourth as well, will tell the tale. If we see shrinking M2 in the 1-week average for weeks 20 and 21 plus a drop in week 22 of around 1.5% or more, then we could be in for a worst-case scenario and we can start preparing even then, scaling out of risk assets and preparing to go short. On the other side of the pendulum, if we see even minor growth in weeks 20 and 21 and then a fall of only 1.2% or less for week 22, we'll be in a best-case scenario and stocks and the economy should probably be safe through the summer and into 2020.

Most likely though we're going to see something in the middle, less conclusive, and we'll have to wait a few more weeks to see where things stand. If that is the case, it'll be dangerous to go short too early. However, on the chance that we have a clear pattern for weeks 20-22, you now know what to do in either case. Here's to hoping for something clear cut and good luck to all.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.