NITTY GRITTY of the ACTIVE vs. PASSIVE INVESTING DEBATE
Does Passive Outperform?
It is common to hear the claim that passive indexing outperforms active management, except for short run flukes. If the claim is correct, then all the time, effort, training and worrying that go into active stock selection is all a fool's errand. This page by J.Norstad is a very good summary of all the arguments PRO passive investing.
But there are a lot of problems with doing a proper analysis of this issue, and academics have provided little guidance. Following are arguments AGAINST the claim, or issues that have prevented any valid conclusion. All Norstad's points are addressed. Notice that proving the assertion wrong does not prove that active investors outperform passive. It means only that it is possible for active investors to outperform.
1) Theoretical vs Actual Returns : The rate of return claimed by the indexers is an index's theoretical total return (as made actionable with an ETF). But investors' realized returns are very different.
- Investors contribute additional savings at market peaks, and withdraw cash at inopportune times, meaning their returns are weighted differently through of the year.
- They try to keep the faith during market crashes, but succumb to fears and exit after sharp sell-offs. Because ETF's are much easier to buy and sell than mutual funds (through the day and at a price you specify) owners may trade away profits - more so than owners of active mutual funds - especially if the MF has a deferred sales charge.
- They know they should not, but cannot resist the urge to market time with sector indexes they think will outperform (also with small cap and value indexes). When indexers do this they call it 'tilting'. When active investors do it they call it 'market-timing'.
- They rebalance between asset classes during the year. They dollar-cost-average their way into positions. Both actions increase costs.
- They delay reinvesting dividends received. Or they spend the dividend. Or they reinvest it in another index ETF.
- The funds they own may have tracking errors against their benchmark index caused by the fund's MER and also by the fund's costs of operations that are not included in the MER. Promises to hedge currency exposure are very poorly executed by many ETFs, not because of costs but because of how the process is executed.
- There is also the cost of advice. For many active mutual funds the cost of advise in included in the trailer-fees paid out of the MER. These reduce their calculated returns. When an indexer pays for advice that cost is ignored in calculations of his rate of return. (Note that management fees should be factored OUT of any comparison of strategy returns. They are a function of a financial product, not a strategy.)
Returns from active management have no theoretical counterpart - although many will have quant models and trading rules. They are always real-life, in-practice returns realized in specified accounts. So, for any fair comparison between strategies the theoretical index returns cannot be used. The actual return of investors labeling themselves 'passive' has to be used.
Academics are starting to measure the returns of indexers. A 2012 paper ("The Dark Side of ETFs") compares the change in returns from one period to the next, between those retail investors who bought their first ETF vs. those who did not. They found that for those portfolios including both indexes and individual stocks, the active portion out-performed benchmark equity indexes, while the passive portion under-performed. The index holdings were a drag on raw returns, Sharpe ratios and alpha. They found that the index's diversification gains were offset by worse market timing.
2) Straw Man Argument : After all their other arguments have been discredited, the indexers always change the subject to "Indexing (passive) is the most appropriate strategy for retail investors". Since most retail investors do not have the time, interest or training to pick stocks, OF COURSE passive indexing is most appropriate. But that is not the issue being discussed. The issue is performance returns, not which strategy is appropriate for whom.
3) The Total Equals the Sum of the Parts: This argument was famously presented by William Sharpe in this article. It argues that "All portfolio returns that beat the market are offset by other portfolios with lower returns ... because the market is the sum of its parts". There are counter-arguments.
- The first is that this is another straw-man argument. No one says ALL active investors outperform. Some do and some don't. Yes, measured together they are only average, but your personal returns will not be the average. You don't care about everyone else. You care about your own returns.
- Sharpe's analysis ignores cash. The most powerful (and easy) way to outperform the market is by retreating to cash during major market crashes. You don't have to time the tops and bottoms perfectly. All you must do is re-enter the market at a level lower than when you exited on the way down. If nothing else this reduces your portfolio"s risk, if not its return. The stocks sold to exit the market are bought by someone, and that someone is included in Sharpe's set of investors. But your cash portfolio is not included in his math.
- Also on the issue of 'cash', in Sharpe's basic argument he models a group of investors owning all the publicly listed shares of some 'market'. As the 'market' goes so goes the the group. But consider what would happen to dividends paid to that group. Sharpe's model has all the outstanding shares already owned. What happens to the cash in his model? The theoretical total return benchmarks presume that distributions are reinvested in the same equities as received. This assumption works for ETFs because they do not own all the outstanding share. But it violates Sharpe's model and math. He uses a model that cannot be duplicated in reality.
- More generally on the issue of 'cash', Sharpe's model ignores cash flowing between the primary and secondary markets that create market cycles. IPOs and insolvencies move cash in and out of the circle of his 'market'.
Academics don't seem to have attempted to quantify the effect of cash on market values and returns, but the NY Stock Exchange has been keeping track of the total $cash balances and total $debt balances in brokerage accounts. The absolute dollar of each is not really meaningful because it grows over time as markets grow, but their net balance is informative. The chart below tracks the net balance moving between net debt in strong markets to net cash in downturns during both the Tech Wreck and the Credit Crunch.
Notice first that investors start liquidating assets to cover debt even before both market peaks. They also restart buying before market bottoms. Investors do not all sell at the bottom, as commonly claimed by the passive camp. It is selling and buying pressure that turns markets at the start, and sustains trends. Notice second where the market is priced when the net debt is $0. In both crises investors got back into the market at prices lower than when they exited. Not only did they escape the depths of the crashes, but they also gained by the difference in exit / entry points. Notice third that it can be years before cash on the sidelines is redeployed. Dividends would be missed. This is a cost but maybe not so important when cash is used to damp volatility instead of permanently owning low-return bonds.
There is an excellent article at Philosophical Economics which argues for a model that puts a flexible ring fence around stocks, bonds and cash - instead of around just stocks. The author uses his model to another purpose, but his analysis supports the point made here - that cash cannot be ignored.
4) Which Math Calculation of Returns : Consider the chosen method of calculating rates of return. There are different methods for different points of view. You must use the correct method for each issue. Take a simple example.
|$500 invested at the beginning of the year. |
50% growth in the first six months resulting in a $750 portfolio. ($500 * 1.5)
20% loss in second half of year resulting in $600 portfolio at year end. ($750 * 0.8)
No one would argue with the claim of a 20% return (600/500 -1) for the year.
But what happens if you add $750 of additional savings to the portfolio at the half-way point? You are not changing any of the investing decisions or their outcomes. The only thing that changes is you double-up the money invested.
|$500 invested at the beginning of the year. |
50% growth in the first six months resulting in a $750 portfolio. ($500 * 1.5)
$750 of additional savings is added June 30 giving a $1,500 portfolio.
20% loss in second half of year resulting in $1,200 portfolio at year end. ($1,500 * 0.8)
If the performance of an individual's portfolio is being measured it is widely presumed that the second half losses should be more heavily weighted in the calculation, because of the larger invested dollars during that period. The IRR methodology would calculate a 5.7% LOSS. This measurement is also called the 'dollar-weighted' rate of return.
If the performance of a mutual fund manager's decisions is being measured it would not be fair to adjust for the added savings because he has no control over those. That decision is made by the unit holders, not him. If you are measuring only his 'active management returns' you would ignore those weightings because his investing decisions were not changed. In fact mutual funds correctly ignore those weightings.
Funds measure the return of an individual unit over the whole year. In this example assume the mutual fund started with 50 units outstanding, each valued at $10. The additional savings at midyear doubled the number of units (an additional 50 units, each valued at $15). The year ended with 100 units, each valued at $12. Resulting in a published 20% return (12/10 -1). This measurement is called the 'time-weighted' rate of return.
The proper choice of measurement depends on your assumptions about what drives the money flows into and out of accounts. If additional savings are being added each month, as saved, then the timings of those additions are not 'investment decisions'. The investor is not trying 'to time the market'. Similarly if cash is periodically withdrawn from an account in retirement, as needed for spending. The 'time-weighted' return is appropriate in these cases.
If you presume that additions and withdrawals are market timing decisions then their effects should be included in the measured rate of return. So the 'dollar-weighted' rate of return is appropriate. Each year there is a report produced by DALBAR that purports to measure the returns of investors in mutual funds - and always concludes that the investors' personal returns are humongously worse than the mutual fund's reported returns. Morningstar produces a similar yearly Mind The Gap report with less extreme conclusions. The underperformance is mostly because they measure 'dollar-weighted' returns. They weight the fund returns for smaller units of time, by the size of the fund. They conclude that investors pile into stocks at market peaks, and sell out at market bottoms.
The false logic of this should be obvious. When investors buy funds, the size of the fund increases. The portfolio manager buys individual stock with his new cash - from other investors in the market place. Fund units are created by moving actual stocks from the freely-traded float into baskets that backstop each fund unit. And vice versa. If fund investors experience terrible returns by getting market timing exactly wrong, then there must be some other group of stock-pickers who experience humongously better returns from owning individual stocks. But all research shows shows the opposite - that all classes of investors make the same timing mistakes. DALBAR's and Morningsstar's conclusions cannot be valid unless they can identify a group of stock-pickers earning the offsetting humongous returns.
5) Mutual Fund Evidence : Most often the only real argument to support the superior returns of passive indexing comes from a comparison of returns between high-fee active mutual funds and low-cost index funds. This argument's use of mutual funds is not valid on many levels.
- The active mutual fund is only a subset of all actively managed portfolios. Even if you prove that active funds under perform indexes, you have not proved that ALL active stock pickers under perform. Mutual funds have a lot of attributes and constraints that do not apply to other active investors. You cannot generalize from the specific to the general.
- Investors in mutual funds face a daunting decision - which fund? - which fund manager?. In reality there is no logical way to make the decision. The only data available are historical results - yet even the funds warn that past results do not predict the future. Unit holders are not making 'investing' decisions. They are deciding who to trust with their money. It cannot be called a rational decision. Their returns will reflect this impossible reality. It is unfair that their choice impacts the measured returns from active investment. It should only be the manager's return that count. He is the only one making 'investment' decisions.
- The people who are sold mutual funds, often freely admit they know nothing about investing, and don't want to know. Their level of misunderstanding can be seen on discussion forums after market down-drafts. They will often post comments disparaging the fund they owned for losing money. They have no conception that the market in total fell, and that their fund might actually be very good since it fell less.
This level of misunderstanding predictably leads to errors like switching funds after a loss, just when the market is due to recover. The returns earned by these people cannot be considered 'returns from investing', much less a proxy for 'returns from active investing'.
- Costs matter. The 2% fees charged by active mutual funds are a headwind not encountered by passive index funds. Fees should be factored out of any comparison of returns because they are not a function of the investing strategy - they are not incurred in the investing process. The issue being argued is the return from different strategies, not the returns from different financial products. Fees vary between financial products, between investors, between countries, between years. They are business overhead costs.
Even when the Management Expense fees of mutual funds are taken out of the comparison, funds are still left with a long list of costs not faced by other active investors. There are registration fees, compliance costs, costs for printing and mailing statements, audit fees, higher transaction costs, etc, etc.
The individual stock picker who enjoys the game, does not consider his time and effort a cost. Supporters of passive investing respond to this argument by sneering at individuals who enjoy the process. They declare that investing SHOULD not be fun and those who enjoy it are pathetically in need of a life. But moral judgements make a poor fall-back argument.
- While it is appropriate to measure the returns of managers without weighting by dollars invested, the reality is that the cash flows to/from unit-holders DO effect the managers' investment decisions. Academics have found evidence for and against this effect.
- E.g. a fund is constrained to a maximum percentage ownership of a company. If that maximum has been reached, new money into the fund cannot be invested in that security.
- E.g. there is an emotional difference between a decision to 'hold' vs. a decision to 'buy'. It may be harder for a manager to add new money into positions he would otherwise be content to 'hold'.
- The fund manager is constrained in his investment decisions by requirements
- to be fully invested at all times.
- to use only one 'style' selection criteria (e.g. value vs. growth).
- to never use derivatives or hedge.
- to never short-sell or buy on margin.
- to hold securities of only one industry sector or geography.
- Fund managers must always hold cash to satisfy daily redemptions. It will generally reduce returns because cash is not productive, although in falling markets this cash will reduce losses. This may be the reason mutual funds are shown to outperform in falling markets. Funds may be using the liquidity of ETFs to replace this non-productive cash. A 2016 paper found that funds with small ETF positions manage cash better and time the market better.
- The fund manager puts his name on the results, but there is a team behind him. His results may depend on a very good analyst. His results may suffer if that analyst quits. No multi-year tracking of fund results can reflect this reality.
- The bigger the mutual fund, the larger each position becomes. When a big fund buys and sells, its orders swamp the market and drive prices up and down. Their returns will be reduced by these market impact costs. The use of computerized trading systems will have helped dilute this effect.
- There is a debatable argument that investing success requires a certain personality with a certain frame of mind. The winners will be free-thinkers with the ego to hold a position in the face of widespread disagreement. This ego is the same thing that allows top athletes to win. Whether testosterone or serotonin, chemicals are produced by winners, and in tern make future success more likely - a virtuous cycle than can turn into a vicious cycle after a year of under-performance.
Think how money managers like Bill Miller must feel after disastrous 2008 results following decades of out-performance. Their egos will be deflated. They will start second-guessing their own decisions. They will almost inevitably continue to under perform because they no longer have self-assurance. But because they are paid obscene compensation they continue in their job until eventually fired.
The retail stock-picker need not suffer that same fate. He too will lose confidence after a year of losses. But he can retreat into ETFs and lick his wounds until he regains his moxy. There is no cash incentive to distort his decisions and cause him to dig the hole deeper.
- Group-think within the confines of the money-management industry will lead to poor investments. You sometimes hear of successful individuals moving to New Hampshire to avoid this contagion of opinions, but mostly institutional money managers are Alpha males who congregate to size each other up.
The idea that the poor performance from mutual funds derives from the attributes of the industry, and not the managers' stock-picking skills, finds support in research by Cohen, Polk and Silli (2010). They identified the individual stocks held with most conviction by fund managers. Those stocks' performance over the next 3 months showed an out performance (relative to a variety of benchmarks) that increased with the level of conviction. The counter-argument is that the high-conviction positions identified by the researchers were simply stocks that had performed well in the past even though given normal weightings at purchase. Momentum would account for their subsequent out-performance.
This begs the question "Why hold all the poor performing stocks if portfolio managers never believed in them to start with?" Fund managers are paid by AUM, the larger the fund the greater their pay. But relentless buying of their top-conviction stocks from new investors' dollars, pushes up prices and lowers returns. Funds may be constrained from holding above some percentage of outstanding shares,so new money will be deployed elsewhere. Highly concentrated portfolios have returns that vary widely from benchmarks against which the manager is judged. They have returns that are more volatile than widely diversified benchmarks.
High conviction portfolios of 20 positions can be professional suicide. Stories abound about fund managers going against the market and losing their job a month before the market turns and their decisions validated. This pressure leads to closet-indexing. The 'theory of tournaments' says that portfolio managers in effective competition with other managers and indexes, react to their own lagging performance by assuming more risk. Their environment changes their investing decisions. The retail investor at home does not feel this pressure.
The only possible conclusion is that mutual funds cannot be used in the determination of the issue "does passive investing outperform active management?" There are just too many non-strategy attributes that distort their results.
6) Pension Fund Evidence : Other research supporting the conclusion that passive investing outperforms, comes from analysis of pension plan returns. Although slightly different, pension plans have most of the same problems as mutual funds. They are only a subset of the actively managed world. Their funds are allocated to money managers just like individual chose mutual funds - with the same propensity to chase past performance. The managers charge fees just like mutual funds charge fees. Those managers' results are benchmarked just like mutual funds with the resulting pressure to conform to an index.
Maybe most importantly their results are similarly public and subject to scrutiny and critique. To play the markets well you must believe in yourself. You need mental toughness to take positions shunned by everyone else and hold through turbulence. The publicity attached to a year of poor performance can wreck havoc on your self-confidence. It is no wonder that future performance suffers when everyone is watching and waiting for you to fail. The retail investor at home need tell no one. We can find all kinds of excuses to make ourselves feel better.
In other ways pension plans differ from mutual funds. Unlike mutual funds that report the manager's performance, pension plans report the performance of the plan itself - using a dollar-weighted rate of return. Then again, managers of pension plan assets do not have to deal with the cash flow problems of mutual funds. The steady cash flow of premiums and benefits are predictable and largely offset each other.
Overall, the drag on returns from these particular structural arrangements make the returns from pension plans not really representative of returns for the active management strategy - just like mutual funds.
7) Academic Research : Academic research seems to actively avoid measuring the returns of retail investors. None of the papers measure portfolio performance the way common sense would dictate (except one, and it does not report the results). Why not measure returns the way normal people would? Ignore transactions. Measure the change in the portfolio's value between the beginning and end of each year. Factor in cashflows. Calculate the time-weighted returns for each year. Then compare to the returns of relevant large index benchmarks.
This would be the way you calculate your own return. This would certainly be the simple method. It is the method that generates meaningful results with the fewest arbitrary modifications. But academics won't use it. Or did they use it and come to conclusions they did not want published? It is unreasonable that this basic research has not been done.
One paper that throws out this measurement as a minor point is "The Dark Side of ETFs". The very high relative weighting of DAX index holdings indicate that German investors benchmark themselves against their home country index. Table VII shows a net Alpha of 4.4% annually for the active portion of accounts relative to the DAX.
Instead, academics create a pure invention. They work from a list of transactions to measure the returns of individual securities, not the portfolio. They ignore cash - which can have the greatest impact on returns and risk. They ignore leverage. They ignore securities held throughout the entire period. They measure monthly (and even daily) rates of return instead of yearly . They assign transaction prices from month-end values and ignore the actual prices. They invent costs to account for the bid-ask spread in their assigned price. They continue to rely on research from the 1990s when transactions costs were ten time higher.
Possibly the largest problem is that they all discount investor returns whenever they feel the return is due to their list of 'risk' factors. The need for risk-adjusting is questioned further in (9) below. Risk-adjusted returns may be of interest to academics, but they have no resonance in the real world. Investors simply do not care about any extra risks in their portfolios. They decide their asset allocation between equities and debt before they divvy up the equity portion. They do not reduce their equity allocation when they find the stocks they own are more risky than the index.
The second largest problem with their research is due to the population used. The population is excessively weighted to very small accounts. These accounts will not fund anyone's retirement. If the investors are in a country with a public pension (like European countries) then this money may, quite properly, be treated like 'play money' using lottery-like gambles. The securities chosen may be because of insider knowledge. Their small size does not really allow for proper diversification. It is not relevant to decide the success (/not) of these portfolios relative to that of someone investing their entire wealth and retirement funding; relative to any wide index. It is especially not relevant to risk-adjust their returns.
There is a recurring theme that gets an honourable mention (but not directly analyzed) in many papers (example). The authors comment that individuals with higher returns in one period seem to continue outperforming in other periods. In other words, some people have 'IT' and other people don't. Since the period being studied is almost always short, it should not be concluded that 'IT' investors will still have 'it' in all markets. But the comments do indicate you may be selling yourself short by simply concluding you don't have 'it' - by indexing - by not even trying.
There is a separate page devoted to the existing academic research. You will find links to the articles and discussion about their results and shortcomings.
8) Consistency : Another straw-man argument is ... "Active investors cannot consistently outperform the benchmark. Managers who outperform in certain types of markets, underperform the other types". No one disagrees because it makes no point. The active investor is free to retreat to passive investing in markets he knows his personality does not suite. It is very common for active investors to index during bull markets but change to stock-picking in recessions. It is the indexer who is constrained by ideology not the active manager.
Why has consistency become a valued attribute? If returns over the business cycle are equal, who cares that in some years stock pickers under performed? Returns will always be 'different' when different assets are owned. Differences from a benchmark do not make a portfolio riskier. Heck, those yearly differences may even out returns over time reducing portfolio volatility.
It is pretty widely accepted that some markets provide better stock-picking opportunities. It is a home truth that active management out-performs during stock crashes. Academics have widened the criteria to include both up and down markets where there is wide dispersion of returns across the market and wide volatility of individual stocks (Gorman, et.al. and von Reibnitz).
9) Benchmarks : Some indexers claim superiority by simply dismissing the reported returns of active investors. They claim these investors do not know how to properly measure returns - either on a year's basis, or multi-year basis. But no proof is presented for this position other than personal anecdotes. Discussion forums provide plenty of evidence that indexers themselves do not measure their returns properly (or at all).
Other stock-pickers' results are rejected by claiming the wrong benchmark was used for comparison. They argue that since the portfolio included (e.g.) foreign stocks or value stock, or small-cap stocks, then the comparison benchmark should also be weighted for these sectors. E.g when Warren Buffett outperforms the S&P Index, they claim "He did not outperform the Value Index and that is what he owns". This argument is wrong because it uses hindsight to determine the benchmark. Buffett is free to own public or private equity, debt, derivatives, commodities, etc. He was never constrained to value stocks. He chose what he owns freely. If he chose value stocks, and value stocks outperform, he deserves the kudos.
Other results are rejected because it is decided the stockpicker's holdings were more risky than the benchmark. A more risky portfolio is 'supposed to' generate higher returns, so its actual returns are dismissed. No academic papers measure the actual returns of investors. Instead they compare a metric they create for investor returns, with a synthetic benchmark that is risk-adjusted. The point of this is to normalize for the different risk levels. Whether you SHOULD normalize for risk is debatable.
- Does the investor care about any extra risks in his portfolio? No. The use of risk-adjusted metrics is valid only if the investor would change his debt-equity asset-allocation to compensate for the added risk of his securities. In real life investors make no AA adjustment. Investors accept the difference in risks as immaterial - whether it is or not. The asset allocation decision is made before deciding which stocks to buy. The stocks chosen are the ones with potential profits, regardless of their supposed risk. He would not make any adjustment to his asset allocation to correct for differences in risk.
- Stock-pickers try to find mis-priced securities. Since these are more likely to be found in smaller stocks not covered by analysts, they end up owning smaller-caps --- because this is where the mis-pricing is, not because they are targeting higher risks. If their strategy is to buy low and sell high, they end up with value stocks - which again are more risky. If they choose to tag along with the market with momentum stocks - these again are more risky. To dismiss their returns because of the strategy chosen, is to guarantee the conclusion that they under-perform.
- Academics discount investors' returns they claim to result from the individual securities' attributes (anomalies) that have been shown to result in generally higher returns. They label these anomalies 'risks' but that is a value-judgement without merit. No one seems to question this. All profits derive from some 'reason'. If you refuse to recognize any profit because it has a 'reason' you are simply refusing to recognize profits. See discussion on the Beauty Pageant page.
- Cash in the account is ignored by academics in their discounting of returns for risk. But it is impossible to measure risk without including cash. Active management's greatest benefit comes from simply exiting the market, into cash, during major downturns. The arrogance of academics pretending they are correctly adjusting for risk, without measuring cash, is unspeakable.
- Any individual's risks come from more than his investments. E.g. If he is an oil worker, and oil stocks are booming, is it appropriate to claim he under-performed the market because he held no oil stocks? No. His portfolio choice was correct. It reduced his risks. But the researcher would have no way to know that, or correctly risk-adjust his returns.
- Discounting returns for risk is only valid when risk is considered a 'bad' thing by the investor. But there is plenty of evidence that many investors actively want more risk than the large-cap index. There is the evidence from low-volatility stocks outperformance (investors wanting risk bid up the risk stocks). There is the existence from the popularity of double and triple leveraged ETFs. There is the existence of ETFs that short the market (where you pay to play, when you could shorted the market for free is you had access to the leverage you wanted).
- A very interesting paper ("Simulated Experience" 2015) found no decrease in satisfaction, after the fact, due to a choice for higher risk and realization of higher variance in returns.
10) Luck vs Skill : Statisticians (like Nassim Taleb of "The Black Swan" fame) dismiss individuals' long-run record of out-performance with the argument: "Those results would be expected from random outcomes, therefore real-life results are due to luck not skill.". E.g. Let 10,000 people flip a coin once. Only those with heads can flip a second time. Only those with heads again can flip a third time. After 5 flips there would still be 313 people in the game with 5 sequential 'wins'. 3% of the original population - and just be chance.
The intuitive error in that logic is shown by simply replacing "the probability of flipping heads 5 times" with "the probability of your teenager's marks being in the top half of the class". The application of the argument is exactly the same.
Situation: Your kid needs grades in the top half of the class to get into college. He comes home one day and announces... |
Kid: I'm not going to study any more. There is no point. My teacher is 'marking to the curve'. Half the kids will do better than average by definition. It is just luck which half I end up in.
Parent: It is not luck, it is talent. Studying changes your personal probabilities. Kids who study most often get top scores. Slackers most often do poorly.
Kid: Nah. They don't do well because they study. Binomial probabilities predict that some kids will repeatedly get good scores just by chance. So their success is due to luck, not talent.
No parent would buy that argument. Did the logic of the statistics argument justify the kid's conclusion? No, because the issue is provable. We can measure time studied and the resulting marks. When the issue is NOT provable (like investing success) then we get seduced by math's authority. But the argument was invalid in one situation and equally invalid in the other.
The argument is invalid because it simply PRESUMES. It does not prove. Just because a statistical model can be created that generates a similar outcome does not prove real-life is random. The only conclusion that can be drawn from the argument is: "IF investment returns were due to chance THEN you would expect some people to outperform over long periods". But the question "ARE investment returns due to chance?" has not even been discussed, much less proved.
A blog by Michael Kitces does a good job of explaining how impossible it is to generate statistical proof. Any excess returns are probably only a percentage or two. These are swamped by the large expected distribution of results (about 20%) due to chance. Statistics is geared to preventing a wrong conclusion that XyZ can out-perform the market. In the process it creates a very high probability that it fails to correctly identify out-performance when it actually exists.
11) Replicatable Systems : Some indexers argue: "Show me a system that will consistently outperform the index ... then we'll see." There is no logic to this argument. This person wants a 'system' with rules that can be followed without subjective evaluation, judgement or thought. OF COURSE this person should be indexing. No one says otherwise. It is common sense that any repetitive and exploitable pattern that can be discovered will be arbitraged away immediately. Stock-picking is not a paint-by-numbers application of rules, it requires skill. It is highly subjective, requiring trade-offs of many different good and bad attributes of each company. The fact that stock-picking cannot be reduced to a set of rules does not prove anything regarding the issue of whether an individual with skill can beat the index.
Burton Malkiel's "A Random Walk Down Wall Street" suffers from this same logical disconnect between the description of reality that few would dispute, and his conclusions. It is like using the argument "University campuses are beautiful therefore you should get a university education." At each chapter's end his conclusion causes you to ask "Where did THAT come from?" In many cases you ask "Didn't he just prove the opposite?"
12) Who Gains When Markets Correct : There has to be a mechanism for the markets to remain 'true' to correct (in hindsight) valuations. Indexing applies capital to stocks in their pre-existing market weightings. It does not change their relative values. Without active managers the markets would become 'wrongly' priced almost immediately. So there is a theoretical gain for investors willing to 'correct' the market. Even while not every active investor will make the right call on 'true' value, in total the correct (in hindsight) active investors must predominate in order to move any stock's price. They will benefit from alpha gains. They reap the benefit from closing the gap between yesterday's valuation and today's new reality.
It is interesting that indexers also use the Efficient Markets Hypothesis to justify passive investing. Their argument is that market prices are always correct and quickly reflect new information - so it is impossible to benefit from mis-pricing. The error in their logic lies in their presumption that the price correction happens by magic, without anyone gaining from it. SOMEBODY must move the price. That person will gain alpha.
The indexers respond with: "But those gains will go to the big boys with millions to spend on analysis and huge trading floors, etc. The retail investor does not have a chance of outgunning them." But that argument contradicts their basic position that returns are a matter of chance only (unless there is access to private information).
13) Chimpanzees Throwing Darts : There is a pervasive myth that you can do just as well as the pros, or the index, by simply taping the stock listings (these used to be printed in the paper each day) on the wall and throwing darts to select your portfolio. The Wall Street Journal put this to a test that was repeated for many years. No matter which camp you are in, you can find support for your position from the results. For a good overview of the competition and resulting analysis see this Investor Home page.
The pros' picks averaged 10.8%, the DJIA 6.8% and the Darts 4.5%. They ignored dividends, but if the average yields for each are added to the score you get Pros = 12%, DJIA = 9.9%, and Darts = 6.8%. At this point all kind of excuses were thought up. You must adjust for risk. You must consider the pricing pressure from publishing the recommendations (in those days recommendations were less ubiquitous). The sample size was too small. Etc.
14) The Zero Sum Benefit : The secondary stock market where retail investors play is a closed system. For most purposes the movement of cash between the primary and secondary markets can be ignored (IPOs and going-private transactions). In aggregate, if some player wins, another player loses. A theoretical argument can be made for the superior returns of stock pickers by balancing the known returns of the other strategies.
- If you accept the measurement of passive returns to exactly equal the theoretical index, then indexers don't weigh-in as either winners or losers on either side of the equilibrium.
- If you accept that active mutual fund investors as well as pension funds underperform the market, then this sub-set of active investors is on the losing side.
- Is there any subset of investors you think guaranteed to underperform? The 25 year old male with no finance training, little cash, using charting, comes to mind.
- Which means some other sub-set of the active investor group must be outperforming the market - us retail investors using more rational strategies.
CONCLUSION: The issue is very difficult to analyze. The most common argument using only mutual funds or pension funds has no merit. The academics have not devised appropriate studies. Nor is it likely they can. Most people believe what they want to believe. What you want to believe derives from your opinion of the Efficient Markets Hypothesis. Specifically, do you consider available information to be CORRECTLY priced into the market, or do you consider all information to be incorporated - but not necessarily correctly.
- If you believe that markets are always correctly priced, then active market timing is a waste of time. But if you believe that markets are driven by greed and fear that frequently overcome rational knowledge, at least temporarily, then you see a place for active managers.
- If you believe that credit and business cycles lead to cyclical corrections, then moving allocations between asset classes makes sense.
- If you believe that the average of all market participants' returns equals the index return, but retail investors never beat the market then active management for them is pointless. If you believe that retail investors have just as much chance of being in the 50% who beat the index, and that some individuals will continue to beat, then you will want to see if you could be in that winning portion.
What you CANNOT conclude is that the issue has been proven one way or the other.
Review of papers on mutual fund performance
Logic analysis of issue by Philosophical Economics
Active switching between funds increases retail investors' returns.
Fund Returns by Degree They Index
Funds Beating the Index Persistantly
Hedge Fund Managers Outperform
Retail investors' herding predicts poor stock performance
Luck vs Skill - Fama, French
Fund Managers' Conviction Picks Outperform