Want to increase returns by 8.5%? Close your online trading account.

A new paper gives strong indication that on average, individual investors are really poor investors. I mean, really bad. Enough that by the end of the day, they’d be better off sitting in cash than trading on their own. It treats them to the sort of analysis which we usually subject mutual fund managers to, and finds that they’re no better, i.e. worst than a passive strategy.

Meyer, Schmoltzi, Stammschulte, Kaelser, Loos and Hackenthal write in “Just Unlucky? A Bootstrapping simulation to measure Skill in Individual Investors’ Investment Performance” :

This paper disentangles skill and luck in individual investors’ investment performance using a four-factor model and apply bootstrapping simulations pioneered in the mutual fund literature to distinguish skill from luck. We use a comprehensive dataset of 8,621 individual investor portfolios from a German online broker, spanning a timeframe from September 2005 to April 2010.

We find that 89% of individual investors exhibit negative skill (α ≤ 0) when measured on a gross basis and 91% when considering returns net of costs and expenses. An individual investor with an average level of risk-taking depicts an investment performance of approximately -7.5% per year for gross returns and of -8.5% per year for net returns. (emphasis mine)

Note that individual investor gross returns were negative. This indicates that the vast majority of individual investors are, for the most part, significantly under-performing a passive strategy, before fees. They are making bad decisions, even before we factor in the cost of executing those decisions. This contrasts to prior findings that gross returns are about the same as the market, and transaction fees reduce the returns.

It’s filled with interesting tidbits, like that the financial crisis in 2008 reduced measured skill, i.e. :

The sample period contains the financial crisis of 2008 and 2009. We find that skill levels of individual investors are less negative in our sample period prior to the crisis, i.e. from September 2005 to the end of 2007.

And the conclusion is fairly strong:

The results are a clear case for passive strategies. Hence, banks, politicians and individual investors might want to reconsider investment strategies and policies to help investors improving the investment skill. In the light of these findings and Campbell’s (2006) call for financial economists to come up with solutions to the investment mistakes of individual investors becomes even more urgent.

The paper has a great review of the literature, including this tidbit on mutual funds:

Generally, there has been some consensus in the fund literature that active investing cannot beat a passive benchmark and that funds rather exhibit underperformance when trying to do so. Hence, the general conclusion is that buy-and-hold appears as the dominant strategy. The most important studies in this respect include Elton et al. (1993), Grinblatt et al. (1995), Gruber (1996) and Carhart (1997). The aforementioned authors find that common factors in stock returns, persistent differences in mutual fund expenses and transaction costs explain nearly all of the predictability in mutual fund returns. Theoretical foundations were also laid by Berk and Green (2004) who argue that, following a long-run equilibrium theory, abnormal fund returns are bid away in competitive markets. The authors show that mutual funds face costs which can be described as an increasing convex function of assets under management. Following this thought, a fund with a positive expected α before costs attracts inflows until its assets under management reach the point where expected α, net of costs, is zero while outflows drive out funds with negative expected α vice versa.

Mental accounting at the pump

Mental Accounting and Gasoline Consumption, by Justine Hastings and Jesse M. Shapiro

“When the price of gasoline increases … the market share of regular gasoline increases while the market share of higher quality gasoline falls.”

While conventional economic theory suggests that decision makers treat a dollar as a dollar no matter how it was earned or is to be spent, in practice some households may engage in “mental accounting” — setting aside special budgets for certain purposes, like food, clothing, or transportation. Households that budget this way may respond to a given income gain, or loss, differently depending upon how it arrives. For example… a household may reduce spending on vehicle-related luxuries more in response to an increase in fuel prices than in response to a comparable loss in financial wealth.

In Mental Accounting and Consumer Choice: Evidence from Commodity Price Shocks (NBER Working Paper No. 18248), Justine Hastings and Jesse Shapiro consider this type of mental accounting with data on purchases of gasoline. Using aggregate data covering 1990 to 2009 and data on purchases at a retailer for 2006 to 2009, the authors find a clear and positive effect of gasoline prices on the market share of regular gasoline, the lowest quality gasoline available. When the price of gasoline increases — typically by similar amounts at all quality levels — the market share of regular gasoline increases while the market share of higher quality gasoline falls.

The extent of this substitution from higher quality gasoline to regular gasoline cannot be explained by income effects alone. During the 2008 financial crisis, for example, the income effect would have predicted an increase in the purchases of regular gasoline and a decrease in the purchases of premium gasoline. In practice, the opposite occurred.

Moreover, the income effects necessary to explain the relationship between gasoline prices and quality choices are extremely high. Households adjusted their mix of gasoline purchases almost 20 times more to a reduction in their buying power because of an increase in gasoline prices than to an equivalent reduction in income from other sources.

Psychological models of decisionmaking may be able to help explain the buying patterns observed in the data. These findings also have interesting implications for retailer behavior — they indicate that consumers will put a higher premium on saving money on gas in high-gas-price times than in low-gas-price times. This implies that retailers will face more intense competition during high-price times. That prediction is borne out in data that shows lower retail margins on gasoline in periods when oil prices are high.

 

Do behavioral interventions have durable effects?

That’s the title of of a new paper by Hunt Allcott and Todd Rogers.

From the abstract:

Interventions to affect repeated behaviors, such as smoking, exercise, or workplace effort, can often have large short-run impacts but uncertain or disappointing long-run effects. We study one part of a large program designed to induce energy conservation, in which home energy reports containing personalized feedback, social comparisons, and energy conservation information are being repeatedly mailed to more than five million households across the United States.

We show that treatment group households reduce electricity use within days of receiving each of their initial few reports, but these immediate responses decay rapidly in the months between reports. As more reports are delivered, the average treatment effect grows but the high-frequency pattern of action and backsliding attenuates. When a randomly-selected group of households has reports discontinued after two years, the effects are much more persistent than they had been between the initial reports, implying that households have formed a new “capital stock” of physical capital or consumption habits. We show how assumptions about long-run persistence can be important enough to change program adoption decisions, and we illustrate how program design that accounts for the capital stock formation process can significantly improve cost effectiveness.

What does this mean? Well, I find it best to view as a stylized graph: 

Every time a mailing is made, electricity consumption drops a bit in response to the feedback about usage. While it has a tendency to tick back up over the next few days, the effect remains. Over many months, a new habit of electricity usage is formed. Once the mailers stop, the new, lower level of consumption remains.

This is the first such paper to look at the long-term, durable influence of such interventions, and it is reassuring to see that (at least in this case), the intervention does not only have a short-run influence on behavior, but also a long-term change in how the individual consumes.

 

Knowledge or Patience?

One of the most provocative questions in behavioral finance is what causes some people to end up with lots of money, and some with very little. Yawn! you say! How is that provocative? It’s provocative causes lead to policy and private market solutions, if you want to treat problems. And the traditionally presumed causes are very different from the ones put forth by behavioral finance. So a recent paper by Justine Hastings and Olivia Mitchell looking at why certain people end up with more or less wealth is provocative because it goes against the common wisdom that the most important lever for improving saving and investment behavior is for financial literacy to be increased. While classic economics says it’s about rationality, behavioral economics says it’s about patience. In their own words:

Two competing explanations for why consumers have trouble with financial decisions are gaining momentum. One is that people are financially illiterate since they lack understanding of simple economic concepts and cannot carry out computations such as computing compound interest, which could cause them to make suboptimal financial decisions. A second is that impatience or present-bias might explain suboptimal financial decisions. That is, some people persistently choose immediate gratification instead of taking advantage of larger long-term payoffs. We use experimental evidence from Chile to explore how these factors appear related to poor financial decisions. Our results show that our measure of impatience is a strong predictor of wealth and investment in health. Financial literacy is also correlated with wealth though it appears to be a weaker predictor of sensitivity to framing in investment decisions. Policymakers interested in enhancing retirement wellbeing would do well to consider the importance of these factors.

To reiterate, patience is a better predictor of health and wealth outcomes than financial literacy. This implies that just teaching people about compound interest and diversification may not improve financial outcomes much. You need to change peoples ability to defer gratification, to be more patient. The earliest evidence I’m aware of about this was in the 1960s. You can read more about delayed gratification and life-long outcomes in this great New Yorker piece.

But they have now replicated the Mischel Marshmallow experiment, so we get to see what patience and self-control look like in action very early on in life (and very cute).

Training for more (or less) than the “main event”

A friend of mine is training for the New York Marathon, and while I’m not planning on running it, I am training with him. This in general has involved increasing the distance we’re running by about 2 miles every other week, and last week we finally hit 12 miles, nearly half a marathon.

Two things strike me about most training plans for marathons:

  • Your training doesn’t usually include distances more than the main event. This is unusual – usually it makes sense to train at levels worse than the main event. if I was training for a 5k, I would definitely run more than 5k, along with time trials over the actual distance. Marathon training plans generally max-out at training runs of about 22 miles however. I guess marathons take so much time and are so hard, you actually want them to be single events, otherwise you’d wear yourself out.
  • I can see the “train for less than the event” effect in my running times. I’ve been running about 10 miles every weekend for the past month. Each time I’ve finished, I’ve felt pretty tired, but good. This Sunday when I ran 12 miles, I could see that I wasn’t used to running more than 10: after 10 miles, my speed dropped off significantly, even though I was on a slight downhill. Likewise, my wife (who usually runs 4 miles) ran 6 miles, and slowed down right at about the 4.5 mile marker.

* Tracked using google MyTracks

This is tough – when training for an extreme event, you might not want to train as much, or even more than the main event because it can be so wearing. But it’s almost guaranteed that when go past your training distance, you’ll find the remainder of your run much more challenging.

When do investors feel regret?

Investor regret: The role of expectation in comparing what is to what might have been

– Wen-Hsien Huang — Marcel Zeelenberg

http://journal.sjdm.org/11/11303/jdm11303.pdf

Investors, like any decision maker, feel regret when they compare the outcome of an investment with what the outcome would have been had they invested differently. We argue and show that this counterfactual comparison process is most likely to take place when the decision maker’s expectations are violated.

We found that decision makers were influenced only by forgone investment outcomes when the realized investment fell short of the expected result. However, when their investments exceeded prior expectations, the effect of foregone investment on regret disappeared.

In addition, Experiment 4 found that individual differences in the need to maximize further moderated the effects of their expectations, such that maximizers always take into account the forgone investment.

What does this tell us, and how can we use it?

  1. We’re  most likely to look for comparisons feel it in falling or sideways markets, when our expectations aren’t being met.  And unfortunately, cash will always be a willing accomplice to our regrets, providing a “what could have been” investment.
  2. Conversely, we’re unlikely to regret underperforming in rising markets, even if the market does better than us.
  3. Some of us (“maximizers”) need to look out for this tendency more than others – and you know this ahead of time by taking a maximizing/satisficing test.

The limits of diversification

Diversification is always good. It’s just limited in how much good it can do.

Diversification is achieved by adding assets into a portfolio which have correlations less than 1 with the portfolio. At its purest level, it reduce risk because not all assets will have the same gain or loss at the same time. By investing in different assets (which all have the same risk and return), we reduce the extreme movements of the portfolio, often in a way which doesn’t reduce the overall return quite as much.

To highlight how this works, let’s take an investment with an expected return of 6, and volatility of 12 (all example assets have these values in this post). We’ll mix the this asset equally with a clone of it, which is completely uncorrelated with the first asset. This reduces the volatility to 8.6 from 11 – a reduction in risk of 22%. The graph below illustrates what happens when we continue to add in more uncorrelated assets exactly like the first. The return stays constant, but the volatility continues to go down. But each time it goes down, it goes down a bit less.

We can look at this directly, and note how the decrease in volatility per increase in assets behaves. The ability to reduce risk falls off quickly, and we appear to hit a limit at a volatility of about 2.7.

 

The examples above give a very simple example of how diversification works, but they are unrealistic in a few ways. First, the assets are completely uncorrelated. It is almost impossible to find assets which are completely uncorrelated in the real world. Let’s run the same analysis again, but this time with a correlation of 0.5 across the assets.

Now we see that the benefits of diversification are strongly related to the (lack of) correlation between assets. If we run this analysis across all levels of correlation, and approximate the minimum level of volatility we can achieve, we get a graph like the one below. From a set of assets which all have an individual volatility of 12, we can reduce portfolio level volatility down to 2.7, but only if we have 20 uncorrelated assets. If the correlations rises to 0.5, our minimum volatility is much higher, at about 8.

So pure diversification – including uncorrelated assets with the same volatility level- always does help reduce volatility, but it the degree to which it helps depends on the correlations. And even completely uncorrelated assets have their limits.

Given that the correlations amongst equity markets worldwide tend to be quite high – about 0.7 – 0.8 – we need to set our expectations about what diversification can achieve realistically.

SP500 Holding Period Heatmap

Illustrating the long-term vs short-term

A while ago the great graphics gurus (sorry) at the NYTimes created a very cool graphic showing the annualized returns of the S&P500 over a long time period:

http://graphics8.nytimes.com/packages/images/newsgraphics/2011/0102-metrics/triangle.png

This was one of the best graphics I’d seen in a while, but there are a number of things I thought could be improved, or used to illustrate another point.

  1. Red doesn’t mean loss. The light red in the picture means a return slightly above inflation. I think it’s misleading to color a real gain as a red loss.
  2. The grey is likewise misleading a bit. Grey is the 20 year median, which equates to 4.1% higher than inflation. That’s not boring and grey – that’s happy and green! It’s possible it’s grey because of how it compares with cash, but cash was not yielding 4% in real terms through much of that period.
  3. Everything is observed in years. But in the short-term, the first few months can really drive a years return.

What is great is that it’s actually fairly easy (ok, it did take me a few hours) to replicate this graph using completely free software. Using the statistical package R, you can literally run this file, and get the graphs below.

First up is just attempting to recreate the graph, which took a bit more manipulation than I expected, but was doable. Things to note:

  1. My data are not exactly the same. The NYTimes graphic adjusts for dividends, average taxes and fees, and inflation. In that respect, their data is superior to mine.
  2. Each cell on my graph represents an average annualized monthly return, even if it represents just one month. Annualizing a monthly return can produce some extreme values.
  3. I inserted a black line at the 1-year holding period.
[Standard disclaimer – this is not investment advice, just for illustrative purposes of what you can do, does not reflect the views of my employer, etc]
What this brings out a bit to me (not discounting point (2) above, is that most of the bright red occurs within the 1 year time period, along the diagonal. 

 

I wanted to focus the graph a bit more on the influence of holding periods on returns, so I rejigged the graph slightly to focus on holding periods. I really like the result:

  1. You can now see how much more volatile 1 year holding periods are.
  2. You can likewise see how there are very few cases when a nominal loss lasts for more than 10 years – almost never. Taxes, inflation and fees will make this graph look worse, but dividends will make it look better.
  3. It focuses me on the long-run. It shows that even if I had invested right before the  worst months of the 1970s, in about 7 years of holding tight I’d be in the same place as someone who’d invested right before the boom years.
  4. I look at this graph, especially every line longer than 15 years (and I’m definitely investing for more than 15 years), and see that it doesn’t matter if I invest now, or a year from now. Over that time frame, there is little risk around choosing when to invest.
  5. We are currently going through quite a tough patch, but the last time we had such a tough patch, it didn’t really affect returns above 12 years. I don’t know what the bottom half of this graph will look like in the future, but the top gives me comfort that it can’t be that bad.

 

 

What do you think?

 

 

 

Financial advice: (potential) market failure edition

The issue of what financial advisers are paid for comes up often, mostly when there is a question about whether they have fulfilled their fiduciary duty in giving advice. There are a variety of different models for how advisers get paid, and all have at least the appearance of some problem with them. There are two main issues:

  1. Agency: Are the advisers incentives aligned with yours – will they be a trustworthy agent? To my knowledge, no firm charges fees which are perfectly incentive aligned. Clients would have to agree to pay based upon how much value advisers have added compared to what the client would have done on their own – a hard counter-factual to assess.
  2. Quality: How do you know you the advice you are getting is worth paying for?

In the UK, the Retail Distribution Review is changing how financial advisers can charge for their advice, and many advisers will be moved to an up-front fee basis, or having to give “on-going” advice, i.e. initiating interactions with the client depending (usually) on what the market does.

To focus on the second issue, many advisers are quite rightly scared that this might hurt their business significantly, but mostly for a psychological reason – people have a real problem paying for up-front financial advice. They aren’t sure of the quality of it (the adviser often knows better than them), and they will not know if the advice was of high quality until a point far in the future.

This opens up a possibility of market failure to (among other things) asymmetric information. A client cannot directly assess an advisers’ quality ahead of time, and thus will discount the value of advice to incorporate the risk it is low quality. Advisers who give high quality advice will know it, and not want to accept low-quality prices for their high-quality advice. As a result, fewer people will pay (up-front) for financial services, because it seems too expensive relative to what they expect from it.

Recently my parents went to a financial adviser to help them plan their transition into retirement. They were a bit shocked by the $800 price tag for about 2 hours of time with the financial adviser – yet this is incredibly cheap by many standards. Wealth management and financial advisers often charge in the region of 0.60% up to 2.5% of assets under management per year, and clients have historically accepted those charges. My parents charge would have amounted to less than 0.005%, yet they were still floored by it.

Independent Financial Advisers already are a well-established market, but it will be interesting to see how people react if they have to pay up front.