A question from a reader on my recent post Me Too!:
I recently ran across Ed Thorp’s “Beat the Market.” I find reasonable his idea that you can take on risks that (almost / essentially) cancel each other out. Find assets that are negatively correlated to buy one long and the other short (he did it with stock warrants in the 60’s but when I started looking into that, well, I’m late to that party, so nevermind).
I’m uncomfortable with shorting anyway, so what about going long in everything and rebalancing when the assets get out of whack? Aren’t a lot of the price movements of various assets (cash, bonds, stocks, real estate, precious metals) the result of money flowing towards or away from that asset? If people are, on net, selling their stocks, to what type of asset are they sending the proceeds? I can’t predict where people will stash their money next, but if I own a little of everything, I’m both hedged against prolonged depression of one asset class and aware of what’s gotten “expensive” and what’s “cheap” now.
Along these same “indexing” lines, what do you think of using ALL the sector ETFs (Vanguard has 11) to index each sector and then rebalance among them as they change in value? How would that application of your portfolio rule 7 differ than when applied to individual stocks? Also, do you think it would be subject to the same / similar danger as everyone else “indexing” as you wrote about above?
[drizzle]My, but there is a lot here. Let me try to unpack this.
Paragraph 1: All of the easy arbitrages are gone or occupied to the level where the risks are fairly priced. Specialists ply those trades now, and for the most part, they earn returns roughly equal to short-term risky debt. They tend to get hurt during financial crises, because at those points in time, fundamental relationships get disturbed because of illiquidity and defaults amid demands for liquidity and safety.
Paragraph 2: First, rebalancing is almost always a good idea, but it presumes the asset classes/subclasses in question is high quality enough that it will mean-revert, and that your time horizon is long enough to benefit from the mean reversion when it happens. Also, it presumes that you aren’t headed for an utter disaster like pre-WWII Germany with hyperinflation. Or confiscation of assets in a variety of ways, etc.
Then again, in really horrible times, no strategy works well, so that is not a criticism of rebalancing — just that it is useful most but not all of the time.
Aren’t a lot of the price movements of various assets [snip] the result of money flowing towards or away from that asset?
Back to the basics. Money does not flow into or out of assets. When a stock trade happens, shares flow from one account to another, and money flows the opposite direction, with the brokers raking off a tiny amount of cash in the process. Prices of assets change based on the relative desire of buyers and sellers to buy or sell shares near the existing prior price level. In a nutshell, that is how secondary markets work.
Then, there is the primary market for assets, which is when they were originally sold to the public. In this case, corporations offer stocks, bonds, etc. to individuals and institutions in what are called initial public offerings [IPOs]. The securities flow from the companies to the accounts of the buyers, and the money flows from the accounts of the buyers to the companies. The selling prices of the assets are typically set by syndicates of investment bankers, who rake off a decent-sized chunk of the money going to the companies. In this case, yes, the amount of money that people are willing to pay for the assets will dictate the initial price, unless the deal is received so poorly that it does not take place. After that, secondary trading starts. (Note: this covers 95%+ of all of the ways that assets get to public markets; there are other ways, but I don’t have time for that now. The same is true for how securities get extinguished, as in the next paragraph.)
The same thing happens in reverse when companies are bought in entire, either fully and partially for cash, and in the process, cease to be publicly traded. The primary and secondary markets complement each other. Corporations and syndicates take pricing cues from the levels securities trade at in the secondary markets in order to price new securities, and buy out existing securities. Value investors often look at primary markets to estimate what the assets of whole companies are worth, and apply those judgments to where they buy and sell in the secondary markets.
Trying to guess where market players will raise their bids for assets in secondary trading is difficult. There are a few hints:
- Valuation: are asset cheap or rich relative to where normalized valuation levels would be for this class of assets?
- Changes in net supply of assets: i.e., the primary markets. Streaks in M&A tend to persist.
- Price momentum: in the short-run (3-12 months), things that rise continue to rise, and vice versa for assets with falling prices.
- Mean-reversion: in the intermediate term (3-5 years), things that currently rise will fall, and vice-versa. This effect is weaker than the momentum effect.
- Changes in operating performance: if you have insight into companies or industries such that you see earnings trends ahead of others, you will have insights into the likely future performance of prices.
All of these effects vary in intensity and reliability, both against each other, and over time. If you own a little of everything, many of these effects become like that of the market, but noisier.
Paragraph 3: If you want to apply rule 7 to a portfolio of sectors, you can do it, but I would probably decrease the trading band from 20% to 10%. Ditto for a portfolio of country index ETFs, but size your trading band relative to volatility, and limit your assets to developed and the largest emerging market countries. With a portfolio of 35 stocks, the 20% band has me trade about 4-5 times a month. With 11 sectors your band should be sized to trade 1-2 times a month. 20 countries, around 3x/month. If it is a taxable account set the taxation method to be sell highest tax cost lots first.
Remember that portfolio rule 7 is meant to be used over longer periods of time — 3 years minimum. There are other rules out there that adjust for volatility and momentum effect that have done better in the past, but those two effects are being more heavily traded on now relative to the past, which may invalidate the analogy from history to the future.
Using portfolio rule 7 overweights smaller companies, industries, sectors, or countries vs larger ones. It will not be as index-like, but it is still a diversified strategy, so it will still be somewhat like an indexed portfolio.
Finally, even if we get to the point where active management outperforms indexing regularly, remember that indexing is still likely to be a decent strategy — the low cost advantage is