Asset allocation

unsplash-image-jUPOXXRNdcA.jpg

How much should you buy of the different asset classes we introduced?

Portfolio theory

This is going to briefly turn technical but bear with me, you won't need to understand every detail. All you need to know is that every asset class has its own future expected:

  • Average return

  • "Risk", expressed as variability of returns over time, and

  • Correlation with other asset classes

Correlation can be positive (when one goes up/down the other one does, too), negative/ inverse (the asset classes tend to move in opposite directions), or close to zero, also called relatively “uncorrelated” (one asset class' price movement has limited impact on the other, and vice versa).

Ideally you want to invest in multiple asset classes that are relatively uncorrelated or even inversely correlated. This includes positively correlated assets that move with different magnitudes (e.g., US and emerging market stocks are positively correlated but when one rises a lot, the other may only rise a bit, and vice versa). See the section on rebalancing below for an idea why.

If you have all three of these pieces of information (expected return, risk, and correlation), you can plug them into various asset pricing models (like the “CAPM” Capital Asset Pricing Model) and figure out with mathematical precision how much you should buy of each asset class to either:

(a) Maximize your return for a given willingness to accept risk (read: expected near-term asset price volatility), OR
(b) Minimize your risk for a desired rate of return

This trade-off is known as the “efficient frontier”:

Efficient frontier

Each dot in this chart represents a portfolio (mix of assets = asset allocation). If you plot all possible portfolios, you will find that some portfolios are “better” than others. Better here means they are closest to the top left of the chart - i.e…

Each dot in this chart represents a portfolio (mix of assets = asset allocation). If you plot all possible portfolios, you will find that some portfolios are “better” than others. Better here means they are closest to the top left of the chart - i.e., they have the highest expected return for the lowest risk (specifically, expected standard deviation in returns). If you trace the line made by the “best” portfolios, you get the “efficient frontier” of optimal asset allocations for a given risk level.

Help me out: If you have a better graphic than the above that I can freely license, please be so kind to drop me a note!

It's nice to be able to get such a mathematically precise answer to what your asset allocation should be, isn’t it?

Trouble is, it all depends on what average return, risk, and correlation you expect for each asset class. And nobody can predict the future. While past asset class behavior and the best available understanding of current market conditions can provide a guide, neither can definitively predict future asset class behavior.

Unfortunately this means there are many asset allocations for your willingness and ability to accept risk for which you can make a reasonable argument. Which one is “right” simply depends on what return, risk, and correlation you personally expect.

You can find a much more detailed discussion on optimal portfolio construction in this Wealthfront white paper.

Rebalancing

Alongside cost averaging, rebalancing is one of those incredible tricks in investing. Here is how it works:

Suppose you are invested in two asset classes, creatively named A and B. Let's also suppose A and B are inversely correlated (when one goes up, the other one goes down). Now say your asset allocation calls for a 50:50 weighting and you have $1,000 invested in each asset class (for a $2,000 total).

Then say in year 1 asset A outperforms and doubles to $2,000 while B underperforms and halves to $500 (for a $2,500 total). You go ahead and rebalance back to your target asset allocation of 50:50 - selling $750 of A and buying $750 of B, resulting in you holding $1,250 of each A and B.

Whenever an asset outperforms its average return for a while, it needs to eventually underperform (or else your expected average return is by definition incorrect). The same is true vice versa. This is called "reversion to the mean".

To illustrate this, imagine that in year 2 asset A and B's fortunes reverse and asset B outperforms and doubles from $1,250 to $2,500 while A underperforms and halves from $1,250 to $625 (for a $3,125 total account balance). Again you rebalance by selling some of B and buying some of A, setting you up for success when the tables turn once again. Each time you rebalance you automatically sell high and buy low, exactly what you want to do as an investor, just without any of the guesswork.

Compare that to if you had not rebalanced after year 1: In year 2, A would have dropped back down from $2,000 to $1,000 and B would have doubled from $500 to $1,000, leaving you back where you started at $2,000 - a whopping $1,125 less than with rebalancing. That's a >50% return for rebalancing in this specific example!

Impact of rebalancing ($)

Excited about rebalancing and getting ready to revisit your portfolio? Check out my posts on How do I optimize taxes? to avoid getting stuck with capital gains taxes you don't need to pay.

Sample portfolios

Alright, let's finally answer the question of how much of which asset class to buy! You already know there isn't one right answer to this and there are a lot of fantastic websites which have done a much better job than I ever could laying out your portfolio options. Here are some of them, before I tell you my personal opinion on the matter below:

Please tell me about others you like in the comments!

Here is what I personally recommend based on what I told you about asset classes:

Option 1: Keep it maximally simple

Buy a globally diversified stock index fund investing in all ~8,000 investable publicly traded companies around the globe, weighted by market capitalization. Simply set up automatic cost averaging into such a fund and then forget about it. One such fund issued for US investors is ETF ticker VT.

If you are investing from elsewhere and are not subject to US taxation, you may prefer to choose a fund that is domiciled in a jurisdiction that is tax efficient for you (e.g., Ireland). You can find relevant Vanguard funds here. Make sure it’s a “Global All-Cap” fund (all countries, all market capitalization sizes). Check out the Bogleheads wiki for a fantastic, detailed discussion of investing from outside the US.

Pros: Extremely simple. Low cost. No guesswork. No need to rebalance. Minimal monitoring or management needed. No bets on any one country, industry, company size, or currency other than how they are currently valued by the market (market cap-weighting). Limited opportunity to get lured by emotions such as fear or greed. No decision paralysis. Just get going.

Cons: No returns through rebalancing. Few opportunities to tax loss or gain harvest. Can’t change the weighing of different market segments.

Option 2: Let someone else keep it simple for you

Buy a target date fund. Target date funds assemble a portfolio of different asset classes, typically including a variety of stock market segments (like US, developed, emerging, dividend, REITs), bonds, and cash. The weighting of the these different asset classes is designed to change along a "glide path" as you near your "target date". The key assumption here is that you will start drawing down your capital towards zero (i.e., withdrawing funds far in excess of the recurring dividends you receive) starting at your target date. This may not be the right approach for you if you are pursuing Financial Freedom by building permanent, “dynastic” capital.

Target date funds usually start out investing you ~90% into some mix of different segments of the global stock market, with the remaining ~10% in bonds. As you get closer to your target date, they shift to a more "conservative" allocation heavy in bonds and cash, ready for withdrawal. However, you already know that I don’t consider most bonds “conservative” at all in 2021.

Target date funds are all the rage in US employer-run retirement accounts as it is easy for employers to justify them as a one-stop shop investment to default employees into.

Pros: Extremely simple. No guesswork. The rebalancing is done for you. Minimal monitoring and management needed. Limited opportunity to get lured by emotions such as fear and greed. No decision paralysis. Just get going.

Cons: Few opportunities to tax loss or gain harvest. You can't change the asset class weightings - you may not want to be in bonds right now, for example (see Asset classes). The assumption that you will switch towards drawing down your capital as you near your target date may not hold for you if you are pursuing Financial Freedom, building permanent capital, and have an effectively infinite investment horizon. Target date funds also charge an extra layer of management fees on top of the underlying funds.

As you can tell, I consider target date funds more of a necessary evil when you don't have better investment options in your employer's 401(k). If you have to invest in them, pick the one with the furthest out target date available, say 2070.

Option 3: Make it more complicated, reap the rewards

Pick a sensible asset allocation, find the best index funds for each asset class, and rebalance whenever any one asset class gets more than a certain percentage (say, 5-10%) off its target allocation. This is what I personally do.

Here is my personal asset allocation and how it compares to current market weights:

Asset classPortfolio weightMarket cap share (2022)Example ETFs (US)Rationale for allocation
US stocks40%60%VTI, SCHB, ITOTUnderweight (50% if you add US REITs)
Developed market stocks25%30%VEA, SCHF, IXUSClose to global market cap if you add non-US REIT stocks
Emerging market stocks20%10%VWO, IMEG, SCHEOverweight as I expect the most growth opportunity in EM and accept their volatility
US REIT stocks10%2%VNQ, SCHH, IYROverweight as I do not own real estate directly; relatively uncorrelated with other stocks
Non-US REIT stocks5%1%VNQISame rationale as for US REITs

In other words, I am personally (a) overweight EM stocks at the expense of US stocks, and (b) overweight REIT stocks at the expense of the markets at large. I also like to keep it simple, allocating in full 5% increments between a min 5% and max 50% weighting. Otherwise I am not betting on any particular industries (beyond REITs), market capitalization sizes (e.g., small cap), or currencies (as I anticipate a globally mobile lifestyle for the foreseeable future). I don’t plan to hold bonds until interest rates return closer to 1980s levels. I used to have access to decent hedge funds via my 401(k) but decided those investments are too black box for me. I am open to adding more angel investments (i.e., startup private equity) as attractive opportunities arise.

If you read the Wealthfront white paper on asset allocation I linked above and played around on their website, you may find my asset allocation is loosely based on their highest risk score portfolios. And yes, I used to be a Wealthfront customer but closed my account (coming soon).

Please let me know your thoughts (and portfolio critique) in the comments! Investing is a continuous learning game for all of us.

Pros: Keep it relatively simple (not dozens of portfolio components, individual stocks, etc.). Low cost. Maximize returns from rebalancing and tax loss harvesting. Limited management needed (buying/ rebalancing only). Pick your own portfolio weightings.

Cons: Not as easy to automate - unless you go with a robo-advisor, which I no longer recommend by default. Requires enough guesswork and engagement to give you the opportunity to get decision paralysis and/or be lured by emotions of fear and greed: Picking your allocation weights yourself may lead you to end up second guessing yourself and changing them. You need to remember to manually buy more shares and figure out how much of each asset class to buy in order to stay close to your asset allocation. Every time you need to make a choice, you may end up attempting to time the market and hurt your returns. I, for one, have at times struggled to invest additional funds in recent years when I had to manually commit to buy at what have been historically high market prices and valuations. Make sure you have lots of emotional fortitude and write an investment contract with yourself before you choose this option!

Please do not blindly copy my personal portfolio choices as they may not be appropriate for your own personal situation.

Please also note our Sprouting Sense terms of use: The content of this blog is for entertainment purposes only. I am not a qualified financial advisor, tax accountant, or lawyer. Please consult authoritative sources, apply your own judgement, and/or seek advice from qualified professionals before making any investment or tax decisions.


OK, we've talked a lot about "asset classes". But what assets are we actually buying and investing in? Glad you asked!

Go back: Asset classes

Previous
Previous

Fund structures

Next
Next

Asset classes