A person looking at their money roll

On Complexity, Diminishing Returns, and Personal Finance (Part 2)

Welcome Back

Welcome to part 2 of my 3 part series examining the ROI of the various major investment decisions that I made over the past decade while building my nest egg. For those of you just joining us, we’re backtesting 5 decisions I made on an investor who starts saving 10% of their $10k per year salary in 1975 and who retires in 2018. Using the real historical stock returns, our saver ends up with nearly $1.3M to retire on. I break down the contribution of each decision to that $1.3M as follows:

  1. Start investing at all: 50%
  2. Invest 10% of my income (vs. 5%): 31%
  3. Start investing at 23 (rather than 30): 5%
  4. Stick with my plan and stay in the market: 14%
  5. Diversification across asset classes: 0%, but reduces risk

Those results are critically important, especially starting early and saving a significant percentage of one’s salary, but they’re pretty basic. Today we’re going to take a look at three more exotic ways to build a retirement portfolio and see how (and if) they contribute to returns in light of how much time they take to achieve.

And now we return to our program, already in progress…

Getting Fancy

So now I was investing worldwide at low cost. These days, you can replicate exactly what I was doing at the time by buying a lifecycle fund from Vanguard or another low-cost provider. These funds will allocate a percentage of your investment to international stocks and the rest to bonds. They will automatically rebalance to the target allocations and they will adjust the allocations towards bonds to reduce your risk as your time horizon gets shorter. All-in-all it’s a great deal. Put your money in every paycheck and you’ll be in pretty good shape.

Of course, I can’t leave well enough alone.

Value Averaging

It turns out that by putting in 10% of my income every paycheck, I was doing a variation of dollar cost averaging (DCA). DCA says that you should put the same amount of money into your investments each time period. By doing so you will buy more of a given investment when its price is lower and less when its price is higher. This sounds like it should increase your returns a bit, but it turns out that in practice investing as much as possible upfront does better on average. This is because on average the stock market goes up and the more money you have invested, the more you get to take advantage of that upward trajectory.

That said, DCA has advantages. It reduces the psychological risk that goes with having to time when to put in your money and then deal with the consequences. Market timing is nearly impossible, so reducing this risk has value. Also, it’s not realistic to put all the money in upfront if you’re investing based on a paycheck… you put the money in as you get it.

Could I do better?

After reading Bernstein’s The Intelligent Asset Allocator, I learned about value averaging, which is like a combination of dollar cost averaging and rebalancing on steroids. Edelton’s Value Averaging provides an in-depth discussion, but the big idea is that instead of putting in a fixed amount of money each time period, you aim to have your investment portfolio have a specific value at each time period. If you are below your target, you put money in. If you are way below your target, you put lots of money in. If you are above your target, you can take money out (and ideally park it in cash for later use when you are below target).

If DCA is “buy low, buy less high,” then value averaging is “buy low, sell high.” It turns out that it’s better than DCA in basically all situations although it still loses to investing a lump sum upfront in many situations. The advantages are that it is possible to do it while having the money trickle in over time and that it reduces the volatility of your investments even more than diversification alone. It also provides very clear targets over time, which I find helps to motivate me to save.

The downside is that it can be complicated to set up. Value Averaging explains how to set up a spreadsheet to do it. I built a yet more complicated spreadsheet to track my diversified investments in even more detail. The spreadsheet took a while to set up and takes more time to maintain. I enjoy it, but you may not.

Let’s see how moving to value averaging affects our results:

8% growth target 10% growth target (with adjustments)
Total invested in 2018 $663,593 $142,993
Total portfolio value $804,188 $816,400

The first thing to notice is that similar to diversification, in this case value averaging resulted in worse results than just putting the money in the SP500 and letting it sit. This is because value averaging basically tethers your return results to the growth rate that you set ahead of time. If you pick a value that’s too low (like 8% for our backtest), you’ll build up a sizeable pile of cash. If you pick a number that’s too high (like for our 10% backtest), you’ll likely have years where you don’t have enough money to keep your portfolio at the target. And how do you pick the growth rate in the first place? Forecasting average market returns over time is very difficult.

One option is to adjust over time. In the second simulation above, I chose to allow these negative years to accumulate until the deficit was large enough that I reset the target portfolio size to be whatever it was at the time. This is admittedly arbitrary, but as you can see above, it doesn’t matter that much. The results of around $800k are very similar. Once you’re on the value path, as long as you can stay close to the path, you have more control over your retirement target.

All said, value averaging is not generally a performance enhancer over the long term. That said, it keeps you on target and by definition dampens portfolio volatility. Further, if you are the kind of person who will make up the difference in bad years by investing even more, it has potential to enhance returns.

Alternatively, If you are looking to keep it simple, look elsewhere.

Taxable Investing

It’s awesome to get a break from the government for investing in your IRA or 401(k), but if you do well like I did, eventually you’ll need to start investing some of your money in a taxable account. I learned that in many situations, exchange traded funds (ETFs) were more tax efficient than mutual funds. They also provide you more options for diversification and if you don’t trade frequently they can cost less. Based on Bernstein’s advice, I began to use ETFs for all my taxable trading, which meant I needed a brokerage account.

I started with Scottrade, but then found that Vanguard was just as cheap (and cheaper if you were trading Vanguard ETFs, which I was often doing). These days, I recommend Interactive Brokers because their trading costs cannot be beat (among other advantages I’ll describe below). The only downside is that their platform is complex and not very user-friendly, which is becoming a theme as we go further down the rabbit hole…

Otherwise, taxable investing is pretty straightforward. Here are a few tips to do it well:

  • Funding
    • If you have a 401(k) match, max it out up to the match limit
    • Then, put as much money as possible into your IRA up to your income limit
    • If you still have money to invest, max out your 401(k) up to your income limit
    • Finally, invest into your taxable account
  • What goes where
    • Bonds, real estate investment trusts (REITs), and commodities are tax-inefficient and should go into your tax-free accounts if possible
    • Stocks, stock ETFs, and tax-managed mutual funds should mostly go into your taxable account
  • How to buy and sell
    • In general, avoid selling things in your taxable account; consider the tax consequences of sales
    • If you have space, consider holding some stocks in your tax free account to facilitate tax-free rebalancing
    • Consider tax loss harvesting to reduce the tax consequences of selling to rebalance your taxable portfolio
  • How to make more space in your tax free accounts if you’re over the income limits
    • You can make taxable contributions to a traditional IRA, which subjects you to more taxes upfront but allows you to rebalance tax free
    • You can do a “backdoor Roth contribution” if you’d rather have the money in a Roth IRA, but be careful because this can increase your near-term tax bill a lot if you do it carelessly

Note that I’m not a tax advisor, so for specifics in your situation, talk to someone who is.

I won’t do a tax simulation here because, surprise, taxes always decrease your returns, often by a staggering amount. That doesn’t mean that you should avoid selling at all costs—sometimes the reward of improving your investment allocation will outweigh the tax consequences. Just be sure to keep taxes in mind once you outgrow all your tax-deferred investment space.

More Diversification (including Factor Investing)

In the previous section, I threw a couple of new investment types at you. Let’s explain those and add a few more that I discovered over the next few years of optimizing and getting progressively more fancy.

By reading Bernstein and Swedroe, I learned that I could get more diversification and lower risk by broadening my investment universe even further:

  • Emerging market stock: stocks from countries with developing markets, like China or Brazil
  • Individual government bonds: diversified bond funds aren’t necessary for U.S. government bonds, which have no appreciable risk of default, so I could buy them directly from the government using Treasury Direct for no fee
  • Real estate: investments in properties held via REITs, which are like index funds of real estate investments
  • Commodities: investments in physical goods held via ETFs or ETNs; I have one for general commodities (like oil and grain) and another for gold

Note that you need to be careful with this diversification. First, stick with low costs (I aim for fees of less than 0.5%) and know that some of these investments will likely have lower returns than stock (especially commodities and bonds) and are most useful for reducing risk, not for enhancing returns. Finally, follow the broad-strokes allocation advice from someone like Bernstein or Swedroe… if you have a silly allocation like 30% in commodities you will increase your risk rather than improve it.

So that’s reducing risk. As I continued reading about hedge funds and the cutting edge of academic finance, I learned that there are also anomalies that may be able to enhance return. According to the efficient market hypothesis, this shouldn’t be possible without also increasing risk. Yet academics and practitioners over the past 25 years have found a number of types of investments that historically have higher returns at the same or lower risk than their alternatives. Better yet, these so-called “factors” have persisted over time, even after their discovery and publication. There seem to be inherent risk and behavioral forces that prevent these factors from being arbitraged away, and we can take advantage of them.

The major factors as discussed in Swedroe’s books are:

  • Size: we already covered this one above, but stocks in small companies (by market capitalization) outperform big companies
  • Value: just like Warren Buffett practices, stocks with a low price relative to their value (by book value, free cash flow, or other measures) do better than high priced “growth” stocks
  • Momentum: stocks that have gone up recently outperform stocks that have gone down recently moving forward (I find this one fascinatingly counter-intuitive)
  • Quality: stocks of profitable, well-run companies outperform stocks in poorly run, distressed companies if you hold price constant
  • Carry: investments that have a higher yield outperform investments with a lower yield

It turns out that momentum seems to be the biggest, most persistent outperformer, followed by value. Size is smaller and quality is less well-defined. Carry mainly applies to non-stock investments. As such, I currently focus on momentum, value, and size. I’m exploring carry.

As I learned about each of these, I was able to find low cost ETFs that divided the world into tranches based on these factors. Altruist FA has great free suggestions for low cost ETFs and mutual funds for size, value, momentum, and geography, which I’ve certainly taken advantage of.

As of right now, my target allocations are as follows:

This is fairly sophisticated stuff, and much more complex than a simple lifecycle fund.

Note that the percentages above don’t matter that much as long as you’re in the ballpark and rebalance regularly. I recommend these broad swath takeaways for a U.S.-based investor for their risky (that is, non-bond) portion of the portfolio:

  • 40-50% U.S. stocks tilted towards value, momentum, and low market capitalization
  • 35-45% non-U.S. stocks, tilted if possible, with extra exposure to emerging markets
  • 5-10% REITs
  • ~5% other lower-return diversifiers like commodities

The percentage of low risk holdings that you have (e.g. bonds and CDs) depends on your investment horizon. The less time you have until you need the money, the more bonds you should be holding. That said… if you’re young and taking conventional investment advice you probably have too many bonds. I don’t have any… but that’s a discussion for part 3.

Let’s wrap up by taking a look at the expected impact of these optimizations:

No Brainer S&P 500 Advanced diversification
Total saved $142,993 $142,993 $142,993
Total value $728,854 $1,283,742 $1,308,205

Here, for my own sanity and a lack of extensive historical data for some of these investment types, I limited my backtest to splitting up the portfolio as follows:

  • 25% large momentum
  • 25% small value
  • 40% international
  • 10% REITs

Like the other backtests, I ignore taxes and transaction costs and rebalance annually. I plug lack of data near the start of the simulation with large U.S., small U.S., and developed markets data respectively.

Here, the fancier diversification does significantly better, about 80%, than the “no brainer” portfolio and its terminal wealth value of over $1.3M is quite good. Surprisingly, it’s only a tiny bit better than just investing in the S&P 500 over that time. Note that, moving forward, we don’t have the value of the hindsight that large U.S. companies specifically will do well. In fact, there is a good chance that they won’t. We do have the factors, however, to guide us towards a similar, potentially better, and almost certainly less volatile result.


We started this post with a fairly straightforward strategy that nets us $1.3M, with the following approximate contribution from each decision:

  1. Start investing at all: 50%
  2. Invest 10% of my income (vs. 5%): 31%
  3. Start investing at 23 (rather than 30): 5%
  4. Stick with my plan and stay in the market: 14%
  5. Diversification across asset classes: 0%, but reduces risk

Now, we’ve considered value averaging, taxable investing, and advanced diversification across more assets and return factors. Was it easily done?

Certainly not:

  • Value averaging adds complexity and may dampen portfolio volatility, but does not generally add returns and may harm them.
  • Taxable investing isn’t really a strategy; it’s just something you have to do if you run out of tax sheltered space. As such, it’s a subset of “invest X% of my income”—the higher you go, the better your result, but taxes will take their cut if you’re not careful.
  • Advanced diversification across assets and especially factors reduces the risk of betting on one particular market segment and usually enhances returns, especially over long time periods; however, the impact of luck on a given sector could be just as large or larger.

Taking everything into consideration, we now rank my investment decisions as follows:

  1. Start investing at all: 50%
  2. Invest 10% of my income (vs. 5%): 31%
    1. Taxable investing: enhances your ability to contribute if you can do it, but less than tax-free investing
  3. Start investing at 23 (rather than 30): 5%
  4. Stick with my plan and stay in the market: 14%
  5. Advanced diversification across assets and factors: 0% to maybe a bit more, but reduces risk
  6. Basic diversification across asset classes: 0% to negative if you include bonds, but reduces risk
  7. Value averaging: 0% or even negative, but can reduce risk and help you plan

This is a bit of a letdown, but it’s also the Pareto principle in action: additional complexity is resulting in diminishing returns.

That’s it for today. If you were reading closely, you may have noticed a teaser that there is something that I only started doing within the past few years that makes a huge difference in my results. It’s the reason that I don’t have any bonds in my portfolio and will be the centerpiece of part 3 of my case study.

Until then, take it easy… but take it.