Why Market Index Investing Works

Why does market index investing work so well?  Most say it’s because you can’t beat the market over the long run, so why try.   That nonsense.  Index investing works by automatically “rebalancing” multiple assets, and by rebalancing more frequently, you can easily beat the market.   Let’s investigate further.

The Components

I’m going to use the Dow Jones Industrial Index for the analysis.  The Dow only has 30 components, making the math easier.  Also, Tobias Preis, Dror Y. Kenett, H. Eugene Stanley, Dirk Helbing & Eshel Ben-Jacob compiled historical data for the Dow from 1939 to 2010 here, eliminating hours of work trying to collect it.  Please note, the returns do not included dividends.

The chart and table below show the performance of the components in the buy and hold strategy.  Since the Dow will replace a stock with a new one at times, this analysis just takes the returns of the newly inserted stock and continues the return stream.  As you can see, a handful of the investments did well, but the majority of them did not.  

Data from Nature.com, Calculations are my own. Dividends not included.

Interestingly, the average annual return of stocks at 8.11% clearly outperforms the Dow’s average annual return of 7.4% (below).  Most people don’t realize this.  Now geometrically (what really matters) the Dow surpasses the individual components’ return by nearly 2% per year.  Remember this fact when reviewing an investment strategy touting individual stock returns.  It’s easy to fool someone into thinking the individual stocks, or a small handful of stocks outperform a market index when in the real world, they probably won’t.

Analyzing Buy and Hold.

We will start by analyzing a buy and hold portfolio, which I equated to game 3 in this prior post.  Our first step is to use the returns from the Dow components above and construct the portfolio.  The portfolio “buys” each stock in the Dow at the start of this time frame, in exactly the same ratio as the Dow in 1940, and then never rebalances any of the investments, letting them all run on their own.  When a component of the Dow is dropped out and replaced with another, we take whatever balance is invested in the stock being dropped, and reinvest it all in the new stock.  In this way, while the components do change over time, we are never rebalancing any money.  The results, as expected, do not keep up with the Dow. 

Data from Nature.com, Calculations are my own.

We see the Dow outperformed the buy and hold of the same stocks by ½%.  Over 71 years, this isn’t a trivial figure, as the Dow ended up 40% higher than the buy and hold portfolio.  So how does the exact same set of return streams end up creating such a different end result?

The Dow is “rebalanced” every time a stock spits or when a stock is replaced from the index.  I’m not going to get into the mechanics of this here, but the effect of these changes in portfolio composition keep the portfolio fairly equal between each stock.  The buy and hold strategy never rebalances, and therefore becomes very skewed by investing in only a handful of stocks as time goes by. 

The following is a chart showing the composition of each component for the buy and hold portfolio.  You can see that over time, the portfolio really consolidates into only a handful of positions.  Three components represent about 50% of the entire portfolio by 2010.  Many of the components are so small, they cease to have any effect on the portfolio at all.   

Data from Nature.com, Calculations are my own.

The lack of balance in the portfolio reduces the portfolio’s diversification, and in turn increases its volatility. As we saw earlier, increased volatility reduces the geometric return of an investment, which is what we see here.  Reduced volatility is partially why rebalancing matters, and why it creates value in an investment portfolio.  Remembering from “The Most Misunderstood Force in the Universe”, over time all investments converge toward the geometric average.  The buy and hold strategy is converging toward the geometric average of stocks, albeit slowly.

The Dow

The Dow isn’t really “rebalanced” in a proper sense.   However it’s construction makes sure a few stocks won’t dominate the portfolio over time which acts similarly to rebalancing.

Below is the composition of the Dow.  You can see it says fairly balanced, although at times, one component can be nearly 10% of the index.  Sharp vertical lines with the chart represent the times when the index “rebalances” through changing components or stock splits. 

Data from Nature.com, Calculations are my own.

So the Dow’s poor version of “rebalancing” by itself creates ½% of extra value per year.  Remember, this is the exact same stream of returns.  One is held forever, the other is kept reasonably equal.  The Dow’s construction and pseudo rebalancing makes sure a few stocks won’t dominate the portfolio over time.  But what would happen if we rebalanced the portfolio strategically?

The Dow, Properly Rebalanced.

The following table shows the results of the Dow components rebalanced to an equal weighting everyday. 

Data from Nature.com, Calculations are my own Dividends not included.

The difference is stark:  80% more wealth than the Dow and 150% more than buy and hold. Nearly a 1% geometric return outperformance per year over the Dow, and 1.5% outperformance over our buy and hold portfolio. 

Data from Nature.com, Calculations are my own. Dividends not included.

Here we created a strategy similar to game #6, the optimal way to approach games of chance with the Dow components.  The Dow (and the S&P500 as well) creates its own “outperformance” over buy and hold through a poor man’s version of game #6.  Really it’s a mix of games #6 and #7.    Remember this is the exact same return stream, for the exact same components of stocks at all times.  The only difference is a proper rebalancing strategy. 

Not Explainable with “Factors”

Normally, academics try and explain outperformance by showing a strategy that favors certain “factors” they believe create better performance.  These are usually represented through the belief that:

  • small companies outperform large ones
  • under-valued stocks outperform over-valued ones
  • stocks with upward price “momentum” show superior performance. 

But none of the above bullets apply to the rebalanced Dow.  The Dow is comprised of only very large stocks.  There isn’t any “size tilt” going on.  We’re not “emphasizing “value stocks” either as the Dow’s price weighting makes this random over time. Momentum tilt?  Really it’s more of opposite, as we constantly sell the highest performers.  Rebalancing creates the extra return, exactly as math says it should.

What Give Rebalancing This Power?

Let go back to a chart from “The Most Misunderstood Force in the Universe”.  Over time the expected compound return falls with each repetition.   If we could, we would prefer to only play the game once.  The same is true for investments in the stock market.  The shorter the time period, the better the results.  In the buy and hold strategy, we ride the investment further and further down the expected compound return path, reducing our returns.      

But when we rebalance, we end the game up front. 

The Dow “rebalances” about every 3 months through splits and composition changes.  The S&P500 “rebalances” less frequently with composition changes, but the 500 different components also means the returns degrade slower with each repetition.  But all indexes “end the game” and then restart the game with a different, slightly more equal composition. Thereby they drive their expected return from the geometric average of stocks toward the higher arithmetic average.

Think of it this way.  You can “play the game” for a month with 30 stocks evenly split.  Some do well, some lose money.  When the month is over, you sell everything.  The game is over.  You only give each stock a month to get eaten away by volatility drag.  So you only go part way down the curve toward the geometric mean.

The average return of your portfolio after one month will be less than the arithmetic mean, but still far above the geometric mean.  For the second month, you play the game again, buying an equal weight of 30 stocks and holding for a month.  Repeat.  The portfolio never gets a chance to degrade over time as you never let it get out of balance, as a buy and hold portfolio does. 

Of course fully selling all stocks and then fully buying them again doesn’t make any sense.  You would simply rebalance.  But it’s much easier to grasp the idea of “ending the game” by imagining selling everything and then buying back in.

Frequent rebalancing is Game #6, the optimum investing strategy, fully come to life.

Relationship Between Rebalancing Frequency and Returns

Below shows the components of the Dow and their returns based on the rebalancing frequency.  This analysis starts with a fully balanced portfolio, so it’s a bit different than the Dow comparisons above. 

Data from Nature.com, Calculations are my own. Dividends not included.

When you don’t rebalance, the returns always converge downward toward the geometric mean.  Rebalancing breaks the cycle.  The more often you rebalance, the further up the curve you go, creating higher investment returns.  Rebalancing every decade is better than never rebalancing.  Yearly is better than every decade.  Monthly is better than yearly.  Daily is better than monthly. Every second is best of all.  Mathematically, you should rebalance as often as you can, until the trading costs become excessive.

Expand to Other Strategies.

Now that you understand rebalancing drives market index returns higher, the proper strategy to beat the market becomes very clear. Just rebalance more often than the market index.

Lets expand this thinking beyond market indexes. How many other investments strategies really only “work” because of the rebalancing effect? I’d say almost all, but we will save that analysis for another day.

22 Replies on “Why Market Index Investing Works

    1. Partially. Smaller companies also usually have higher arithmetic returns, which in a portfolio that rebalances, will usually help as well.

  1. Do the realization of short-term capital gains have any effect on frequent rebalancing? I [mostly] understand the math and principles, but I wonder about the practical effect of this. Am I misunderstanding or is this not an issue?

    1. Yes, short term capital gains taxes will be higher in this strategy than a simply buy and hold. Its a very hard thing to do in a universal way because everyone has different tax rates. I’ve tried to calculate taxes with current tax rates in the back test, and don’t think they are overly burdensome though, I will discuss this in a future post.

      That said taxes are not an issue in an IRA.

  2. The capital gains tax is hugely important in any decision to rebalance. Furthermore, if you are building wealth to leave to your children, it may make more sense to buy, hold, and pass the asset on with a new cost basis that essentially negate any capital gains tax. Rebalancing makes sense mathematically, but must account for taxation before you can make a convincing case.. Bill Launder

  3. Great post, but I’m still trying to wrap my head around this — Is this specific to the Dow with it’s unusual price based weighting? A market cap weighted index like S&P 500 won’t ever rebalance and does become distorted over time based on relative performance of its components. What explains the outperformance (relative to the average stock picker) of these market cap weighted indexes? Shouldn’t your average stock picker (assuming no skill, effectively random picks) who diversifies evenly between 50 or so names and rebalances regularly outperform the S&P 500 too?

    1. If I interpret it correctly, He saying its better to rebalance against cash than other stocks. This is certainly true in some environments. I don’t think its true in most markets though as the returns from stocks historically has been high enough to not provide an advantage to owning cash.

  4. Comparing this to the games discussed in an earlier post, the game is completely random. We know the probability of each outcome is 50%. So this assumes that stock movements are random? They may be random in the very short-term, but individual stocks also often go up because their business fundamentals, growth prospects, etc. improved. So wouldn’t we want to continue owning the stock at the higher price as it appreciates (or as it has “hot hands”)? Otherwise, if we sell it, does that mean we believe the price increase is random and it we expect it to decrease soon to compensate? Likewise, a falling stock may be a faltering business, and buying more would hurt our returns.

    1. No we don’t sell some of it because we expect it to decrease soon. We sell some of it because its now too high a percentage of the portfolio to maximize geometric growth, assuming all else is equal. Now if you now think after it went up in price it should return more in the future than you did before then you might not sell any of it, you might even buy more.

  5. I find it interesting that in the entire discussion of why index investing works you don’t mention one of its super powers: its imbedded survivorship bias. In the real world, individual equities can go to zero and this is a particularly acute reality for the champions of Buy and Hold. For indexes, these problematic securities drop out and are replaced. I also appreciate that frequent rebalancing may keep Frogger on the screen at all times so these considerations are perhaps academic. Love the blog, working through it.

  6. Love the website and the thought-provoking ideas. Have you looked at threshold rebalancing vs. time-based? In theory, it would a) capture some of the momentum premium (i.e. stocks that go up continue to go up, at least for a while); and b) reduce transaction costs, as you only buy/sell stocks that are x% out of balance.

    For example, on the Permanent Portfolio, waiting to sell/buy until one of the components increases to 35% or drops to 15% appears to outperform annually or monthly rebalanced portfolios over long periods of time.

  7. Such an amazing blog. Thanks so much for it!

    Now that you have proven rebalancing as frequently as possible is the better way to go, What about the optimal number of stocks to hold in the portfolio?

    1. Optimally, it’s the most you can, but transaction costs go up with each additional stock, as would tax implications, and the complexity of the calculations if your trying to be very accurate. You then get into a tradeoff between frequency of rebalancing and the number of stocks held (more stocks, have to be rebalanced less frequently). I don’t know where the limit is from that perspective, but there certainly will be a limit. It’s an interesting idea to look into.

      1. I found a couple of posts related to this you might find interesting. It’s about crypto but It could easily apply to stocks/bonds.https://hackernoon.com/crypto-users-who-diversify-perform-better-new-research-ebf775d348ddhttps://hackernoon.com/rebalance-vs-hodl-a-technical-analysis-6f341b0db9cdhttps://hackernoon.com/portfolio-diversity-a-technical-analysis-c2c49f4d3a77https://hackernoon.com/portfolio-rebalancing-for-cryptocurrency-7a129a968ff4

        This one its pretty interesting too. Diversifying into the top 6 or 7 cryptos by market cap way outperforms just holding BTC or ETH alone.https://www.reddit.com/r/CryptoCurrency/comments/ofu03b/which_is_better_buying_top_coins_vs_buying_btc/

        Given this and your amazing blog I decided to make a python script to automatically rebalance my crypto portfolio (using a threshold instead of time).
        You can check it out here if you are interested:https://github.com/burdo3417/rebalanced-crypto-portfolio

        I need to find the right % per coins, but it’s hard because they are all extremely correlated.

        1. Thanks for sharing!

          Do you hold your crypto assets on an exchange? I’ve been thinking about diversifying my BTC holding but was a bit wary of the idea that storing all the my crypto holdings on a third-party exchange. Would love to learn about your thoughts on this.

          1. Hi! Glad you liked it.
            I don’t like having my assets on an exchange neither. I keep 80% of each of my cryptos in a hardware wallet and 20% on an exchange; so I can rebalance when needed.
            Let me know if I can help with anything else!

  8. Great post – pretty much every factor strategy in the literature (small caps, value, momentum, etc.) use rebalancing and equal weighting as part of the strategy. Haven’t seen any discussion in the literature about how much of the factor premium comes from EW + rebal.

Leave a Reply

Your email address will not be published.