Here’s a follow-up to my post about the return gap of retail investors due to poor timing. For every seller of a stock share, there is a buyer. Therefore, if the timing of retail investors is reliably a little worse than average, we also know that someone else is on the other side of all those trades. Is there a group of non-retail investors that is reliably making money off the “dumb money” trades of retail investors?

Larry Swedroe digs into this question in his Advisor Perspectives article The Suckers at the Investment Table:

New research confirms that institutional investors, such as mutual funds, outperform the market before fees, and they do so at the expense of retail investors. That is bad news for retail investors and for investors in active mutual funds, who underperform after fees.

The research finds that the stocks and bonds individual investors buy go on to underperform and the ones they sell go on to outperform – demonstrating that retail investors are “dumb money.”

Unfortunately for fund investors, the same large body of evidence demonstrates that while mutual funds generate gross alpha, their total expenses exceed gross alpha, resulting in negative alphas for their investors.

If on average, an actively-managed mutual fund generates 0.7% of gross alpha, but after you subtract the expense ratio and trading costs which add up to nearly 1%, the net alpha is still negative. The active manager is the winner, taking all of the alpha for themselves in the form of relentless fees taken as a percentage of the entire asset base. The retail investor/customer still loses out. An fairer fee structure would be to take a larger percentage, but of the alpha only.

People will continue to argue about this, but I’m not surprised to see that these studies found alpha. It’s just much, much harder to do than most people think, and that’s exactly why you almost never see a fee structure based on alpha (thought they do exist). Even Charlie Munger, who is famous for his stock-picking skills and disagreement against the “hard” form of Efficient Market Theory, only says that the top 3% to 4% of professional investment managers will outperform (source):

I think it is roughly right that the market is efficient, which makes it very hard to beat merely by being an intelligent investor. But I don’t think it’s totally efficient at all. And the difference between being totally efficient and somewhat efficient leaves an enormous opportunity for people like us to get these unusual records. It’s efficient enough, so it’s hard to have a great investment record. But it’s by no means impossible. Nor is it something that only a very few people can do. The top three or four percent of the investment management world will do fine.

In the end, costs always matter. If you find a genius to pick stocks but they cost more than they help, then you still lose. The only actively-managed mutual funds that I have seriously considered buying are from Vanguard, which improves the odds with substantially lower expense ratios and a history of investor-friendly practices. As a DIY individual investor buying index funds, you can keep your head down and “grind out” reliably above-average returns over time due to the rock-bottom costs. (There, I fit in my own poker reference!) Even as a DIY individual stock investor, as at least I understand what I own and don’t have to pay a 1% management fee every year.

Periodically, Vanguard publishes asset class return forecasts for the next 10 years. Here is their most recent one for September 2021. To their credit, they have also published a recent follow-up post tracking both those forecasts alongside the actual returns in retrospect. This is a good lesson on the difficulty of any sort of short-term market prediction, even after allowing yourself 10 years and wide error bars.

Below is a chart tracking their forecasts for the future 10-year average annualized returns of US stocks. Eyeballing their chart, for 2010-2020 their range of confidence was somewhere between roughly 5% and 10% annually, for a median around 7.5%. This accounts for their model’s 25th percentile to 75th percentage range of possible outcomes. This is a pretty big range! $100,000 times 5% annualized returns after 10 years is $163,000. $100,000 times 10% annualized returns after 10 years is $259,000.

Even accounting for that wide range, their forecast for US stocks was off. While the curve looks vaguely similar, US stocks did significantly better than their forecast:

Again, even accounting for the huge range of guesses, their forecast for international stocks was also off. Global non-US stocks did significantly worse than their forecast:

Vanguard’s original chart focuses on their 60/40 portfolio, which happens to look a lot better. Why? Their 60/40 portfolio consists of 36% US equities, 24% global ex-US equities, 28% US bonds, and 12% ex-US bonds. For one, future bond returns are much more simple to predict than stock returns. Your current 10-year yield is going to be pretty close to your eventual 10-year return. In addition, their US equities forecast was really wrong in one direction (too low), while their international equity forecast was wrong in the other direction (too high), so they tended to offset each other. Is this diversification in action? Certainly, yes, but also luck in my opinion. Both could have also been wrong in the same direction.

Should we just ignore this stuff completely then? I keep thinking back to this illuminating chart comparing the contributions of earnings growth, dividends, and P/E ratio changes to the total return of the S&P 500. Earnings growth and dividends have been pretty consistent for over 70 years, but the overall swings in return have been mostly caused by P/E ratio expansion and contraction.

Consider the analogy that P/E ratio expansion and contraction behaves like a rubber band. It can stretch pretty far, much farther than you might expect, but as you keep stretching it, the stronger it will eventually want to come back. But you never really know how far it can stretch, or when it will snap back. Forecasts can be wrong for a long, long time. You have to balance knowing that the run will end eventually, but not knowing when. Someone will always be right in retrospect.

If only we could focus solely on the earnings growth and dividends. Those are what really matter in the long run. This is the behavioral benefit of remembering that dividends are a share of profits being distributed to you as a business owner. Even if prices on a screen are dropping, the businesses are still working hard, making profits, reinvesting some for more earnings growth, and sending some of it to you as cash.

Just over 45 years ago on August 31st, 1976, the late Jack Bogle started the first index mutual fund at Vanguard. It nearly didn’t get off the ground, garnering only 7% of the initial funding goal – it wasn’t even enough to buy shares of all the stocks in the S&P 500! Read Bogle’s own take from exactly 10 years ago at this 9/3/2011 WSJ article to better appreciate it took determination and stubbornness to make this happen.

Industry leaders mocked “Bogle’s folly”, wondering aloud why anyone would voluntarily agree to be “just average”. Well, Bogle had common sense and simple math on his side. He knew that over time, his fund was guaranteed to be above average due to it’s low costs. Some active mutual funds would outperform for a while, sure, but would there be reliable persistence in those superior returns? It turns out, very little. These days, people worry more about too much index fund investment.

Over time, more and more investors have realized the power of low costs. They are running away from high-cost funds. Morningstar just released it 2020 U.S. Fund Fee Study (free with registration). From the Executive Summary:

The average expense ratio paid by fund investors is half of what it was two decades ago. Between 2000 and 2020, the asset-weighted average fee fell to 0.41% from 0.93%. Investors have saved billions as a result.

(Thank you, Mr. Bogle.)

This chart shows the new investments flows into the cheapest 20% of funds (blue) against the remaining 80% (red) over the last 20 years:

Morningstar research has demonstrated that fees are a reliable predictor of future returns. Low-cost funds generally have greater odds of surviving and outperforming their more-expensive peers. […] Since 2000, net flows into funds charging fees that rank within the cheapest 20% of their Morningstar Category group have trended higher. Flows for the remaining 80% of funds have been negative in nine of the past 10 years.

Vanguard and it’s dirt-cheap index clones are winning. If you look closer, it’s the really low-cost funds that are gathering the most new investment. These are mostly the big names that have started competing directly with Vanguard on cost – iShares, Schwab, Fidelity, SPDR.

Of the $445 billion that flowed into the cheapest 20% of funds and share classes in 2020, nearly all of it went into the cheapest of the cheap, as 93% of net new money flowed into the least costly 5% of all funds.

Investors voted with their money. Follow this trend and continue to effect change with your investment choices. Look for a low-cost fund option in your 401k, and ask why if you don’t see it.

According to Multpl.com, the P/E ratio of the S&P 500 is now 35, and the dividend yield is only 1.3%. The all-time low dividend yield was 1.11% back in August 2020. Even if you consider stock buybacks, the earnings yield is less than 3%! That means if corporations all distributed every penny of their profits as dividends, it still wouldn’t be higher than 3%. Before we go any further, I’m not advocating market timing, as people were saying that the S&P 500 was “overvalued” back in 2015 when the P/E ratio was 25 and the dividend yield was 2%. Nobody truly knows what will happen to prices in the short-term.

Even if the P/E ratio seems a lot higher now than the historical average, what has that actually meant? The Morningstar article How Much Has the Market Benefited from Investor Optimism? examines how much of the historical return of the S&P 500 from 1976 through March 2021 was from P/E expansion.

In January 1976, the P/E ratio was only 11.8. In March 2021, the P/E ratio was 31.5. That seems like a huge difference, and over that 45-year time period, it did add 2.2% to the overall historical average annual return. But we also got 3% from earning growth, and another 2.75% from dividends, for a total return of ~8% above inflation. 8% real return!

In a way, this is somewhat comforting, as if you look at the long-term, a shrinking P/E ratio won’t completely destroy your retirement by itself. Instead of adding 2%, it might subtract 2%.

Looking ahead, if you assume a generous 4% from earnings growth, 0% from a constant P/E ratio, and 1.3% from dividends, that’s roughly a 5% future real return. But if the P/E ratio goes back even partly back to historical averages, that will be closer to a 4% real return. The problem is that bonds are giving us 0% real return at best, so I’m sticking with owning productive businesses.

The numbers on my brokerage statements keep going up so perhaps I shouldn’t complain, but I sure hope the earnings start to catch up to the prices soon (as some predict). I like the idea of the P/E ratio going down due to higher earnings rather than lower prices!

My contrarian thought of the day? I feel that the retirement planning industry downplays the role of luck. Life is not a as certain as the smooth exponential curves that they show you. Perhaps the statistically optimal bet is to jump a bit early and hope for the best, while having a backup plan for the worst. You might just win an extra 10 years of freedom.

If you tinker with portfolio survival calculators like FireCalc and cFIREsim that model hundreds of possible paths, you may notice that the “failure” paths usually happen when a bear market occurs soon after you retire. If you keep spending when a portfolio is down, it may never recover.

Even if you have the same portfolio size, same withdrawals, and the same average returns, having the bad years occur upfront can lead to failure while having the bad years at the end can lead to success. This is known as sequence of returns risk.

Retiring in 2000 with a 4% withdrawal rate: Warning! 🚨🚨🚨 In 2021, most people happily accept that the stock market just goes up and up. However, every so often there will be a “lost decade”. If you retired in 2000 with a portfolio invested in the S&P 500 and used a 4% withdrawal rate (increasing each year by 3% for inflation), here’s how that would have looked like (yellow line):

Retiring in 2010 with a 4% withdrawal rate: More money than you started with. 💰💰💰 If you have solid returns upfront, then you gained a decade of priceless freedom! For retirees of the “Class of 2011”, consider that their portfolio is likely larger today in 2021 even after a decade of withdrawals.

Retiring in 2021? Crystal ball is cloudy. If you are in the retirement “Class of 2021”, many predictions call for another lost decade. Yet, even if the next 10 years have poor returns, better times may be right around the corner. From this article by Davis Advisors:

Though frustrating, stretches of disappointing results for the market are not unprecedented. History shows however, that these difficult stretches have been followed by periods of recovery. Why? Because lower prices increase future returns. – Christopher Davis

This article was written in 2012, and it turns out that Davis was right. As of Q2 2021, the trailing 10-year annual return of the S&P 500 is over 12% annualized. Here is a chart showing the subsequent 10-year performance after each past “lost decade of stock returns”.

Surviving the first 10 years of retirement. The lesson here is to avoid taking out big withdrawals during a stock market slump drop during the first 10 years, so that it can benefit from the rebound of the next 10 years. At the same time, you don’t want give up the chance of 10 extra years of freedom. Therefore, perhaps the best bet is to retire when you have a reached your chosen savings target (for example, 25 times annual expenses), but also maintain a detailed backup plan during the first 10 years. Here are some things you might include in that plan:

• Plan ahead for way that you can temporarily cut back on spending if you need to. Big to small. For example, plan to move to a lower-cost city, country, or housing option.
• Identify non-essential assets that you will sell if you need to. Vacation property, etc.
• Maintain employment opportunities in your current career field. Go back to part-time, freelance, consulting, etc.
• Have alternative employment plans in a different career field to create supplemental income.

(By “plan”, I mean written out on a piece of paper. This improves the clarity of your thinking.)

The most powerful way to counter “sequence of returns risk” is variable withdrawals – a fancy term for the brilliant idea of not taking out as much money from your portfolio when it is getting beaten down. But the first 10 years is the most important, and the first 10 years is probably the easiest to go back to the workforce in a limited capacity.

Bottom line. Deciding when to stop working can be a difficult, personality-driven decision, but one option is to retiring with 95-98% odds of success with a practical backup plan, rather than waiting several more years and reaching 99.5% odds of success. Accept that luck matters (and also that you might have to go back to work). However, you also might gain extra priceless years of freedom. Life is never 100% certain anyway.