Can Supply and Demand Predict the Stock Market?
Working in England in the 14th century was difficult to say the least. Between the harsh conditions and low wages, you also had to fight against the occasional disease outbreak and a repressive government. As the Black Death spread across England, killing anywhere from 30%-50% of the population, King Edward III issued a decree to freeze wages and prevent workers from asking for more. Anton Howes summarized what happened next:
When I look at this data I notice two things. First, the average allocation to equities tends to decline during market crises (e.g., DotCom, GFC, COVID, etc.). This makes sense as we would expect equities to make up a smaller portion of an investor's portfolio as prices fall.
The second thing I notice is that the allocation to equities has been above its long-term average for the vast majority of the last decade. Some of this is due to the strong equity market over this time period (2015-2025) with relatively few crashes. However, it also suggests that investors are allocating a larger portion of their portfolio to equities than they have in the past.
Unfortunately, if you look at the historical data, a higher allocation to equities usually meant lower future returns. The plot below highlights the negative relationship between the investor allocation to equities (from the AAII survey) and the annualized total real return of the S&P 500 over the next 10 years:
In general, when the average allocation to equities was high, future returns were lower, and when the allocation was low, future returns were higher.
This relationship held from 1987 until the GFC, but has since started to break down. Take the most recent data for example—the average allocation to equities in 2015 and the returns over the next decade (i.e., until 2025). This was a period of high allocation to equities and high future returns.
You can see this in the chart below which shows the average allocation to equities for each month in 2015 and the annualized total real return of U.S. equities through the same month in 2025:
If the most recent data suggests that owning more equities doesn't translate to lower future returns, then maybe the AAII data isn't as predictive as it used to be.
For example, when I analyzed the AAII data back in 2018, I discovered a tactical model that would've outperformed a Buy and Hold approach with far less volatility. Here's how that model (which I've called the AAII AvgEquityShare model) worked:
As the plague still raged, in 1349 Edward III issued an emergency ordinance to try and contain the economic fallout. Even though half the population died, their gold and silver coins survived, so that there was suddenly twice as much coinage in circulation per head. And so one of the immediate effects was for the price of everything, including both goods and services, to rapidly rise. This rapid inflation, brought on as it was by so many people dying, inevitably led to higher wages being demanded for all kinds of work. “Seeing the necessity of masters and great scarcity of servants”, the ordinance explained, workers now found themselves able to pick and choose who they worked for, and to hold out for much higher wages than before.While the ordinance (with its threat of imprisonment) was successful in preventing wages from rising too rapidly, the market found an alternative solution to this problem—non-monetary compensation. Once again from Howes:
Although the statute stipulated the amount of cash that workers could be paid, both with and without providing them with food and drink, it didn’t say anything about the quality or even the quantity of that food or drink. So instead of giving them bread made of rye, barley or beans, perhaps with a slice of old salted bacon to sweeten the deal, employers now had to provide their workers with only the best-quality wheaten bread, and with freshly-cooked meat still warm from the pot. Instead of providing them with mere water to quench their thirst, employers now had to give them the freshest of ales. As one contemporary complained, the servants were now demanding “to be better fed than those who hired them”.Ultimately, it was supply and demand that determined how workers were compensated, not royal decree. This example illustrates how market forces find a way to triumph over human intervention. But is the same thing true in the stock market? Can supply and demand tell us anything about future stock returns? Some data suggests so. The American Association of Individual Investors (AAII) has tracked aggregate investor allocations to stocks, bonds, and cash going back to 1987. Over this time period, individual investors allocated 62% of their portfolios, on average, to equities. You can see this in the chart below, which plots the aggregate investor allocation to equities over time (in black) along with the long-term historical average of this measure, 62% (in gray):
When I look at this data I notice two things. First, the average allocation to equities tends to decline during market crises (e.g., DotCom, GFC, COVID, etc.). This makes sense as we would expect equities to make up a smaller portion of an investor's portfolio as prices fall.
The second thing I notice is that the allocation to equities has been above its long-term average for the vast majority of the last decade. Some of this is due to the strong equity market over this time period (2015-2025) with relatively few crashes. However, it also suggests that investors are allocating a larger portion of their portfolio to equities than they have in the past.
Unfortunately, if you look at the historical data, a higher allocation to equities usually meant lower future returns. The plot below highlights the negative relationship between the investor allocation to equities (from the AAII survey) and the annualized total real return of the S&P 500 over the next 10 years:
In general, when the average allocation to equities was high, future returns were lower, and when the allocation was low, future returns were higher.
This relationship held from 1987 until the GFC, but has since started to break down. Take the most recent data for example—the average allocation to equities in 2015 and the returns over the next decade (i.e., until 2025). This was a period of high allocation to equities and high future returns.
You can see this in the chart below which shows the average allocation to equities for each month in 2015 and the annualized total real return of U.S. equities through the same month in 2025:
If the most recent data suggests that owning more equities doesn't translate to lower future returns, then maybe the AAII data isn't as predictive as it used to be.
For example, when I analyzed the AAII data back in 2018, I discovered a tactical model that would've outperformed a Buy and Hold approach with far less volatility. Here's how that model (which I've called the AAII AvgEquityShare model) worked:
- Start by being fully invested in U.S. stocks (i.e., S&P 500 index fund).
- When the average equity allocation goes above 70%, sell all your stocks and move into bonds (i.e., 5-year Treasuries).
- Stay in bonds until the average equity allocation drops below 50%, then sell all your bonds and move back into stocks.
- Repeat steps 2-3 until rich.
- DotCom Bubble: The model sold stocks in September 1996 and didn't re-buy them until October 2002.
- Great Financial Crisis: The model sold stocks in May 2006 and didn't re-buy them until November 2008.
I do not recommend changing your investment strategy based on the following information: The AAII AvgEquityShare model sold out of stocks in January 2018 after being fully invested every single month since November 2008.Yes, the model that called both the DotCom Bubble and the GFC signaled to get out of stocks back in January 2018. Thankfully, I didn't listen. If I had, I'd still be in bonds today and would've missed a 196% increase in the S&P 500 since. Supply and demand is useful for understanding (and possibly predicting) many markets. But it doesn't work as well with stocks anymore. Why? Because the nature of demand has fundamentally changed. Today, investors are far less price-sensitive than they were in the past. With the rise of automatic 401(k) contributions and passive indexing, billions of dollars flow into the markets every month regardless of valuation. This is the premise behind Just Keep Buying. Did this relentless bid for stocks finally break the relationship between high equity allocations and low future returns? Or was it the common knowledge that we share about markets? After all, everyone "knows" that the stock market "always" recovers. And everyone knows that everyone knows that. So, if everyone knows to buy the dip, how does a dip ever sustain itself? I don't know, and I don't look forward to finding out. Until then, happy investing and thank you for reading. If you liked this post, consider signing up for my newsletter. This is post 507. Any code I have related to this post can be found here with the same numbering: https://github.com/nmaggiulli/of-dollars-and-data