2019-06-20

AI & Asset Allocation, Part 2

Second part of an opinion in three parts


Snapshot of Edwin Lefevre in 1907
"You remember Dickson G. Watts' story about the man who was so nervous that a friend asked him what was the matter.
'I can't sleep,' answered the nervous one.
'Why not?' asked the friend.
'I am carrying so much cotton that I can't sleep thinking about it. It is wearing me out. What can I do?'
'Sell down to the sleeping point,' answered the friend."
-- Edwin Lefevre in 1923


Part 1  |  Part 2  |  Part 3  |  Appendix 1  |  Appendix 2  >

In Part 1, we considered and opined about: 
1. The strenuous life on the beach of your kind of long-term investor.
2. It’s hard to consistently beat the investment performance of the S&P 500 in the long term.
3. Questioning the value of AI (artificial intelligence) for tax-loss harvesting, 
4. Doubtful value of AI’s daily portfolio rebalancing.
5. AI portfolio choice, of specific investments, not proven, yet.
6. Costuming assessment of investor’s risk preferences to look like AI.
7. We don’t know AI’s adequacy for the next big crash.
8. Bond values fall when interest rates rise, and vice versa.
9. You probably already own some bond equivalents, like Social Security.

Continuing from Part 1 with our discussion of bonds...

Have a good, clear reason for buying bonds, otherwise you forfeit the higher return available in, say, stocks. Verify your good, clear reason with someone whose financial advice you respect. (If you wouldn’t ask a hairdresser if you need a haircut, then why ask a person who sells bonds if you need to buy bonds? The person who urges you to buy bonds from them may have much wisdom and may be the most honest person you know, but if they get a commission on the sale, then let's bear in mind their conflict of interest.) 

about ETFs…

An ETF (Exchange Traded Fund) is a pool of assets in the care of a manager who issues share certificates in the pool and who buys and sells investments for the pool consistent with the prospectus (every ETF has a prospectus, a document stating the organization of the ETF). Each certificate represents a proportional claim on the assets of the pool. The shares trade on an exchange, such as the NYSE (New York Stock Exchange) or NASDAQ, so you can buy or sell them like stock certificates. In this article, we consider three kinds of ETFs:

Bond index ETFs - The manager buys and sells bonds for the pool to match the performance of an independently determined index. There exist many indexes, but we are interested in indexes of broad classes of government or high grade corporate bonds.

Sector index ETFs - The pools contain common stocks managed to match the performance of an independently determined index of a particular industry, country or other narrow category. For instance, you can buy sector index ETFs to match the performance of the health care, telecommunications, real estate, and petroleum industries, and of the publicly traded companies in Switzerland, China, Japan, United Kingdom, Eurozone, and Chile, and of many other countries and industries.

Broad Stock index ETFs - The pools contain common stocks managed to match the performance of an index reasonably defined to specify “the entire market”. The classic, oldest examples are S&P500 index funds VOO and SPY, which match the 500 largest companies traded in the US. There are broad stock index ETFs that attempt to match all the publicly traded companies in the world, and more specialized in the largest companies of particular countries (indistinguishable from country sector index ETFs), and in large-capitalization (big) companies, small-capitalization (little) companies, and mid-cap (medium sized) companies. Also, we find “growth” stock variations, which contain the cross-section of companies with low book-to-price ratios, and “vaue” index ETFs, which contain companies with high book-to-price ratios.

On the other hand, as long-term investors, we have little interest in ETFs which aren’t managed to match the performance of an index, but which are managed according to the manager’s judgement, also called “actively-managed” or sometimes “hedge” funds.

See more definitions in Appendix 2.
If you decide to buy bonds, then buy bond index ETFs (Exchange Traded Funds), plain vanilla bond index funds invested in US govt securities or high-grade US corporate bonds, such as ticker symbols SCHZ, ILTB, SPTL or VTIP. Use ETFs, because selecting individual bonds is tedious and arcane, requires frequent attention to your holdings to reinvest on maturity of a bond, and because only those who trade hundreds of thousands of dollars in each trade can justify the time spent on this work. With ETFs, you avoid those concerns. 
If your concern about whether you have enough invested in bonds and cash keeps you awake at night, then buy enough bond ETFs that you can sleep at night. Cash has nearly no investment return, but it won’t lose its face value, so many regard it as extremely low risk. If you are more than 30 years old, then keep enough cash to pay your living expenses for 3 to 6 months, and invest the rest of your cash in something.

That leaves us with stocks...

A definition ... when talking about stocks, "cap" is "capitalization", which is the number of shares multiplied by the price.

To get significant rewards, you must take risks. Use diversification of investments and long-term holding periods to moderate the risks you must take.

For investments outside of the US: Mr. Mark Hulbert has found that returns on US versus ex-US stock indexes ran, historically, in long 5+ year episodes where one is higher than the other. Generally, there is no reason to buy ex-US stock ETFs at this time.

Your portfolio should consist of ETFs of broadly diversified stock index funds and perhaps some bond funds (see warning about bonds above), unless you want to invest in some individual stocks (see Part 3), in which case, reduce the stock ETFs to 95% of your stock investments, or to 90%, if you must. 

If you are shy about risking fluctuating prices, then buy a S&P500-indexed large cap fund like VOO or SPY to avoid extreme fluctuations. If getting as much increase in value as you can reasonably expect interests you more, then get a large cap growth index fund like VUG or a small cap growth index fund like SLYG. Higher rates of growth come with increasingly large expectations of fluctuations of prices over the next 2 or 3 years, but if you are going for the long 5+ years horizon, then your long 5+ years holding period mitigates the price fluctuation risk. The relative return advantage of a growth fund, such as VUG, is significant over a ten-year period, compared with its vanilla siblings, such as VOO. Small cap growth funds show only small return advantage versus large cap growth variants.

So, for broadly diversified stock index ETFs, in order from higher rates of value increase (return) to lower:
1. Small cap growth -- expect highest price fluctuations, highest rates of growth in value.
2. Mid cap growth.
3. Large cap growth. 
4. Large cap S&P 500 -- expect lower price fluctuations, lower rate of growth in value.

You can also buy value index ETFs, instead of "blend" growth+value (plain vanilla like the S&P500) index ETFs or growth index ETFs. I don't think the value index ETFs are worth the price fluctuation risk forgone, but if you are worried, put 2%-points of your stock portfolio in bond ETFs and buy growth index ETFs.



My recommendation for stock index ETFs: Go for growth. 

Exception: if you expect high (a large part of your personal budget) and rising medical expenses, then you might like to have an investment that would increase in value with your medical expenses, and then you might substitute health-and-pharma sector funds (for example XBI, IHF or VHT) for a large portion (up to 1/4) of your investment in broad stock index funds.

For choosing broad stock index funds, I have found these metrics relevant: 

“When forced to choose, I will not trade even a night’s
sleep for the chance of extra profits.” -- Warren Buffett
1a. 10-yr growth of price (credibility weighted, giving more credibility to all-funds average versus specific fund observation for funds younger than 10 years, and if you can't do that calculation, then ignore this metric, and ignore funds younger than 10 years),
1b. 10-yr growth of price (ignore funds younger than 10 years),

2. Book value divided by price (inverse of price divided by book). Book value divided by price correlates negatively with future performance, that is, a low score is better, as in golf.

Sensibility and Prudence: It's sensible and prudent and low in effort to put 100% of your stock portfolio into ETFs of broadly diversified stock index funds. You may give up a percent point or two of returns in many years, but the 100% solution is a lot easier and a lot less labor than trying to get just a little more return. If you choose this 100% solution, then you can make your portfolio choices, order your trades, and never sell. (And if it allows you sleep better, put a portion of it in bond funds, but see warning about bonds above). Then go outside and play, and stop reading here.

On placing small bets on sector funds and individual stocks… see Part 3.



Disclosure

I own shares of VTIP, XBI, VOO, VUG, SLYG and VHT.





I want to thank my proofreader, 
whose corrections and comments were 
indispensible.




Part 1  |  Part 2  |  Part 3  |  Appendix 1  |  Appendix 2  >



Images, notes and sources

See complete list in Part 1.










Wild foxglove. Gig Harbor, Washington, USA, June 2019.
Image by Daniel Brockman.