The Demographics of Stock Market Returns Part II

The approach described in the last post, coming from an interesting and well-written research note from people at the Federal Reserve Bank of San Francisco tracks relatively well the general trends of price to earnings in the USA since the 1950s, which suggests that the prediction of the price to earnings going forward are also precise, and these predictions suggest a bear market for the next decade.

A word on identification errors and cognitive bias
I discussed a few issues related to this approach in the last post, one of which is the use of long-dated time series to measure statistical parameters that are supposed to track markets, as well as the omission of many potential variables that could make the bear market turn into a huge bull, and vice versa.

Before I continue, let me quote Jesse Felder, in March 2015:

“I think there’s a very simple explanation for the high stock market valuations since 1990: demographics. From 1981–2000, the baby boom generation came into their peak earning and investing years. Is it just coincidence that during that very same time we witnessed the largest stock market valuation bubble in history? No. In fact, there is a statistically significant correlation between demographic shifts like this and stock market valuations.”

This is convincing, and maybe he is right. Certainly the fact that “the baby boom generation came into their peak earning and investing years” did play a role in boosting stocks. But it may also be a case of identification error and information / cognitive bias.

Here is how an identification error happens: I have a strong belief that nothing special changed in the economy since whatever period in the past, and I also strongly believe that demographics is a powerful force that explains most of stock market bull and bear markets (which, again, MAY be true). I then observe a phenomenon (bull market), and I match it to a “cause” that fits my beliefs (demographics).

The “Observed phenomenon” is the 1990s bull market. Suppose “Potential cause 1” = “the baby boom generation came into their peak earning and investing years.”

What other “potential causes for the bull market since 1990” could we think of that would be 1) realistic and 2) potential major drivers of stock markets? These would be “Potential cause 2”, “Potential cause 3”, and so on. Here is a non-exhaustive list that I come up with very quickly, off the top of my head:

1. The fall of the Berlin Wall in 1989, the end of the Soviet Era, and the flocking of global capital to the one country representing prosperity, the USA.
2. The general globalization of capital markets and the emergence of global supply chains, facilitated by unprecedented communication and computing technology advancements, in both scope and pace
3. A significant increase in productivity growth relative to the 2 prior decades, thus pumping up profits and future growth potential.
4. The North American Free Trade Agreement that came in 2 steps, in 1989 with Canada, and 1994 with Mexico, causing lots of growth, lower unemployment, decreasing production costs for companies listed on stock markets, and lots of capital influx to US markets.
5. A low-but-positive-and-stable inflation decade, thus decreasing inflation risk and interest rate risk, and pumping up capital markets.
6. The “global savings glut”, with Asian savings flooding global and US markets, driving real interest rates down further and pumping up asset prices even more.
7. China joining the WTO and providing extra purchasing power to Western consumers through lower priced items and lower global inflation and interest rates.
8. And most importantly: the COMBO and inter-connections of many of these elements… not counting that I must be missing many.

Do these elements “justify” the disconnected-from-reality price to earnings of the dotcom bubble of 2000 or the credit and housing mania leading up to 2008? NO! Bubbles are explained by mass psychology effects, “Animal Spirits”, easy credit, and leverage. But these elements DO provide “a few” other reasons for a bull market, beyond demographics, don’t you think?

So which one of these (and potentially others) was the main driver of the bull market? Well, without a detailed look into data and details, your guess is as good as mine, and guessing sucks.

If you are heavily into ego-driven and/or emotional investment decisions, you will simply pick the “thing” that you “feel” is the main driver, and you will build a very convincing “story” around this, and the demography thing will be your main storyline for financial advice… and maybe you will be lucky once or twice due to “other factors” going your way and reinforcing your demography story, but maybe “other factors” will drive stock markets in the opposite direction from your beloved demography-based story, and then you will spend a tremendous amount of psychological processing power explaining and justifying why your calls are off by 5 miles.

Is a secular bear market upon us?
That is what the demographics suggest, because proportionally more retired people means proportionally less buyers per seller, hence a fall in stock prices, all else equal. Again, I am NOT saying demography is not important. In fact, I think it IS.

The issue is that there are “a few other things” (AHEM!) that could change that story, just like the previous discussion we had to “explain” the bull market of the 1990s and beyond. Again, here are “just a few” items I can quickly come up with, off the top of my head, in 5 minutes… I am probably omitting many…

1. A small more-than-expected increase in the 65+ labor participation along with a small, more-than-expected increase in the proportion of 65+ wanting stock market returns, even with the risk. Note that this amounts to a change of preferences in a sub-group of the 65+.
2. Less people retiring than what is expected, along with a quest for stock market returns by some sub-group of these older workers planning to work until they are 70 or 75.
3. A slightly lower-than-expected unemployment rate for key age groups, making extra savings magically appear in markets and adding to capital asset demand.
4. A slightly higher-than-expected employment-to-total population ratio, meaning a less-than-expected dependency ratio, causing more production than anticipated, hence more income, and more asset demand.
5. A decrease in capital controls in China that makes a good chunk of (stratospheric) Chinese savings flock to US markets.
6. ​Fundamental innovations and global growth that allocates part of the extra income to US markets.
7. The lagged effect of the fastest fall in global poverty in History that brings global savings partially to US markets.
8. The growth of India (which will have an increasing working age population relative to total population) that becomes a “new China”, supplying more capital to global markets. Note: although Africa has a long way to go for growth and development, the same demographic story applies for that part of the world as well.
9. Latinos and other foreign-born people coming to the US (who will represent a growing share of the total population AND who have a totally different demographic pattern) may have greater stock investment preferences than the typical American resident, thus causing a small increase in the proportion of stock buyers in the 25-65 age group.
10. Less growth in the stock supply in stock markets (let’s not forget supply in the supply and demand story!).
11. Higher profits caused by stronger-than-expected global growth and a rising global class of consumers buying goods and services from US multinationals.
12. Regulation changes domestically and internationally that allow a greater proportion of stocks relative to bonds in portfolio mixes of financial institutions.
13. An increase in the willingness to pay for stock in a financial context where there are very low returns in safer asset classes (sounds familiar?).
14. And last but not least… a combination of some or all of these!! Just a small effect from a combination of all these could more-than-counteract the demographic aspects and totally flip the story and give plenty of fuel for a bull market story.

Perhaps you are reading some of these and you don’t “buy the story”, you say to yourself “yeah right”… that is OK. It doesn’t matter – I have the same reaction to some of these myself. But if you are reading in good faith and not desperately clinging to the doom scenario of the demography story, surely you admit that some of these are possible, and you see that just a few of these combined could bring a never-seen-before bull market that takes everyone by surprize and kills the “doom by demographics” story.

Am I calling a major and historical bull market? NO! Because all of these elements and others, as well as the demographic outlook could ALSO bring the worst market conditions ever seen! You just have to keep the whole picture in mind, assign realistic weights to the different elements, and not get all caught up in information bias and cognitive capture that blurs judgement.

Does the Japanese experience prove the demographic story?
Maybe. Aging is certainly one element, and the working age population will decrease in absolute terms for Japan for many years to come. But we would need a full picture of what affected Japan since the 1990s: zombie banks, excessively low inflation, a lack of foreign investment in Japanese markets, a domestic preference for bonds, zero immigration, etc. The list is long, and any conclusion that makes a direct link between the demographics of Japan and the stock market again oversimplifies the picture.

Let’s have fun with actual scenarios!
Just for fun, let’s try a few numbers, OK? We will build the “base case” with actual numbers from the BLS, the Fed, and the US Census Bureau. Then we will fiddle just a LITTLE bit with a few fractions in a realistic way, and build a huge bull case… Ready for a few laughs? OK go!

The basic story is that we need to build the “buy group” and the “sell group”, then evaluate the evolution of the buyers-per-seller ratio over the 2015-2030 period.

Here is the base case:

We take the total population aged 25-64 and apply the current aggregate average stock holding ratio (the proportion of the population that holds stock directly OR indirectly, through mutual funds, pension funds, etc.), and this will be the “buy group.”

We take the 65+ group and apply a very high seller ratio, to signify that the majority of 65+ are sellers of stock. This is the “sell group.”

We then calculate the ratio of buyers-per-seller in 2015 and in 2030. If it drops dramatically: SELL, SELL, SELL! If it rises dramatically, BUY, BUY, BUY.

I hear your objection: buyers are really in the 35-55 group, and sellers and really in the 60-70 group. I hear you, and I have good or bad news (I don’t know how you will take it)… it doesn’t change a THING! I tried every possible combination and the SAME “small parameter fragility” flips a bear story to a bull story and vice versa! So please just go along… after all, you go along with others that just huff and puff and provide next to nothing in rigorous thinking ability, bring up one metric, and scare the living hell out of us! So I think I deserve that small 2 cents of trust and respect. OK let’s have fun…​​This is the base case: a drop in the ratio of buyers-per-seller from 1.96 to 1.33… SELL! SELL! SELL!

Now suppose that the US population realizes that it is getting dismal returns due to its excessive risk aversion and its logical consequence: a love affair with bonds. Suppose the ratio of buyers goes up from 50% to 55% (this is realistic), and the ratio of sellers within the 65+ group falls from 90% to 85% (this is also realistic), because a fraction of the people in that group hold their stock longer or even pass it on in heritage. The buyers-per-seller ratio goes to from 196 to 1.55… still a “bear market story”, but not as bad.

Now let’s add 3 very realistic combos to this story:

1. The US Census Bureau makes a SMALL mistake in projections (but well within standard errors, so very realistic) and the 25-64 group is 51% of total pop in 2030 (instead of 49.5%) and the 65+ group is 19% (instead of 20.6%);
2. Add just 2% extra of the global rising middle class in emerging markets of at LEAST 900 million (very conservative estimation) in the next 15 years that want to put money in US stock markets directly themselves or (more realistically) through the recently-matured financial institutions in those markets;
3. The low-yield obtained in all other markets (bonds, real estate, etc) forces a higher percentage of the population to invest in the stock market to get some returns, and the percentage of the 25-64 that want stock goes from 50% to 60%.

These small and realistic changes make the buyers-per-seller ratio go from 1,96 now to 2.08 in 15 years: a mini bull market is emerging! BUY BUY BUY! OUPS!

Then add just a few very realistic small twists in a combination of global markets, financial market expansion and institutional development in emerging markets, rising savings from increasing GDP per capita in emerging markets, global QE, a small return to optimism in US prosperity, a small realistic decrease in the growth rate of stock supply (totally possible), and all other possible combinations of the listed points above (and surely many more) and you very easily and realistically get a strong INCREASE in the buyers-per-seller metric and a convincing bull market storyline. I built one realistic case in which the buyers-per-seller ratio doubled between now and 2030… a huge bull call!

The same realistic approach could be taken to build the most bearish story you ever heard, and I would sell a million ebooks by “rigorously” calling the end of stock markets…

Hopefully, you find me ridiculous right now! The point is that you can’t pull out one metric and call the boogeyman. It just isn’t that easy, sorry…

“Make things as simple as possible, but not simpler” – Albert Einstein

Conclusion
What to make of all this? As I said in my last post, the bears have plenty of material to convincingly defend their cases: government debt, Eurozone problems and risk (this one is actually quite real by the way!), the excessive spending habits of lazy North Americans, the “fall of China”, deflation, demographics, pollution, the dying middle class in industrialized economies, and on and on and on.

Bulls also have plenty of ammo: global growth and rising standards of living, decreasing poverty in huge chunks of global population, and a rapidly rising global middle class, technology and innovation in health, transport, energy, and communications that are going to surprize many in the next 20 years, improving structural policies and data quality for better macro policy and investment decisions, and on and on and on.

The takeaway is that doom scenarios are just as possible as bull calls. The capitalist system has surprized many sceptics in the past by its resilience and capacity to create growth, innovation, value, and wealth, as well as reduce global poverty and expand accessibility to capital markets to more people.

Stick to medium term projections (0-4 years) with a long run goal of maximizing returns according to your risk-return preference, always keep the “structural aspects” in mind, and allocate your investments based on most-probable outcomes, and rigorous research by competent people who don’t wear doom glasses nor rosy coloured ones.

Pascal Bedard
pbedard@yourpersonal.com