Success Rate Of Retail Investors

Investment is not only a skill but also a temperament. That's why many value investors say you either get it or you don't.

Being a skill means you have to practice, practice and practice. Temperament means you have to excercise control over emotions, anxiety, impatience and impartiality etc. But a lot of people focus on the skill rather than temperament. That's why you can have fundamental houses like Sequoia that couldn't make it after the founder passed on. Or hedge fund houses like Brevan Howard once the founder step down (and now returned). It is a real issue with Berkshire in my opinion.

Why is it surprising that retails don't do as well? I won't be surprised that those who cut their own hair during this COVID-19 period don't do it well either If you don't want to spend effort then it is better to just buy index funds. That's the gist of Buffett's recommendation though it has been twisted by ETF salesman. Like Karlmarx mentioned, there is no simple way, just like any disciplines. There are many grade 1 and 2 pianists but if you want to be grade 8, then it takes different effort. And you can't really pinpoint what is the most important part of the process of music making.

More surprisingly is why institutional investing don't do well. Besides fees and charges and cash drag in a inflationary environment, there is behavioral and institutional imperitives at work. There is conflict because job safety becomes a real concern because you are evaluated monthly, weekly or even daily. There is no incentive to make a long term contrarian call, similar behaviour to the Taiwan political landscape. The same person with MSCI as benchmark or SIBOR as benchmark will behave very differently. A hedge fund with 20% profit sharing will be more keen on short term profit rather than long term performances as the risk reward is skewed (hence a lot of hedge funds are actually beta funds, max out profit & risk to get max 20% profit sharing and if fail then just close and move on). As size increases, as Buffett mentioned many times, the amount to move the needle increases and you are stuck with the index stocks to generate small alpha... before fees. On AGGREGATE it is statistically impossible for funds to outperform the market. That's the point that Buffett is making with the $1m bet with the fund of hedge fund. One layer of fees is difficult enough...

But on individual basis or even good fund manager basis, it is not impossible. Actually Buffett already addressed this in his "masterpiece" of "Superinvestors of Graham & Doddsville". To follow up with what Karlmarx mentioned, I think there are odds; there's no "luck" per se. You can jaywalk 10k times and get hit once (the concept of 万一). That's not really luck, probably we shouldn't jaywalk in the first place, though there are known unknowns and unknown unknowns. Similarly what is the mortality (ie risk with ref to details like age and pre-existing condition) vs probability of contracting COVID (with ref to safe-distancing and mask wearing, hygiene etc) that policy makers will need to assess just like investors. It may sound morbid but similarly they have to assess whether it is worth $93billion. I have seen fundamental investors and speculative traders make money consistently that it's difficult to attribute to "consistent luck", albeit they have vastly different timeframes and techniques.

Good thing about ETF is that it does weed out the average long fund that are basically index trackers. The bad thing about ETF is that it is entering a euphoric stage just like every new products will go thorugh... to an extreme. Index compilers don't do fundamental research: they do structural research in terms of market cap, free float, ease of access etc. That means the large will become larger and the small will be neglected. That's basically what has happened in the past 10 years and also why value and small cap has been lagging. There is serious asset misallocation here. The capital market is no longer efficient in helping new and innovative companies raise capital. You have to be a start up unicorn to be noticed.

Over the long run, as many rational investors has realised, it is like a plane on autopilot guided by numbers-based indices as maps. With systematic and computer trading added in, the flash crashes are already symptoms of the underlying issues. There will be a day when the newest 737 MAX appear. Nobody knows when a bubble will pop but in my opinion most astuste observers are able to see when there is a bubble.

People have too much faith in tech and numbers. And I have been following / covering tech for almost 30 years, only to be called non progressive Only sailors fear the sea.

(16-06-2020, 11:56 AM)specuvestor Wrote: Good thing about ETF is that it does weed out the average long fund that are basically index trackers. The bad thing about ETF is that it is entering a euphoric stage just like every new products will go thorugh... to an extreme. Index compilers don't do fundamental research: they do structural research in terms of market cap, free float, ease of access etc. That means the large will become larger and the small will be neglected. That's basically what has happened in the past 10 years and also why value and small cap has been lagging. There is serious asset misallocation here. The capital market is no longer efficient in helping new and innovative companies raise capital. You have to be a start up unicorn to be noticed.

Over the long run, as many rational investors has realised, it is like a plane on autopilot guided by numbers-based indices as maps. With systematic and computer trading added in, the flash crashes are already symptoms of the underlying issues. There will be a day when the newest 737 MAX appear. Nobody knows when a bubble will pop but in my opinion most astuste observer is able to see when there is a bubble.

Hi specuvestor, what do you think of ETFs that invest based on some quant style? For example, one that focuses on value. That would give the advantage of wide diversification without the problem of ignoring fundamentals totally. And there will be no succession problems.

(16-06-2020, 04:57 PM)gzbkel Wrote: Hi specuvestor, what do you think of ETFs that invest based on some quant style? For example, one that focuses on value. That would give the advantage of wide diversification without the problem of ignoring fundamentals totally. And there will be no succession problems.

Since this topic is on retail investors, I would not recommend these ETFs to them. Though it sounds like a great idea, it defeats the purpose of having an ETF in the first place, which is low cost. An ETF that is based on a quant underlying index will incur higher expenses due to the need to consistently rebalance the underlying.

@gzbkel,
I am currently reading John Bogle's Stay The Course.

The book probably will answer your question on investing in quant-style ETFs. Besides low cost, John Bogle also stated and demonstrated that there is a difference between index funds and ETFs. And then within ETFs, the various sorts (value based, growth based, simple index etc). In fact, John Bogle who started the first low cost index fund, rejected the idea of an ETF when it was first mooted to him. If John Bogle started quant based ETFs, low cost would be guaranteed but it wouldn't be what he recommended for the retail person.

I will leave it up to you to read his book and answer your question yourself. I wouldn't deny you of the joy and privilege of discovering the answer yourself.

You need a correct temperament, even when investing in ETFs (regardless of type).

(16-06-2020, 06:24 PM)ghchua Wrote:

(16-06-2020, 04:57 PM)gzbkel Wrote: Hi specuvestor, what do you think of ETFs that invest based on some quant style? For example, one that focuses on value. That would give the advantage of wide diversification without the problem of ignoring fundamentals totally. And there will be no succession problems.

Since this topic is on retail investors, I would not recommend these ETFs to them. Though it sounds like a great idea, it defeats the purpose of having an ETF in the first place, which is low cost. An ETF that is based on a quant underlying index will incur higher expenses due to the need to consistently rebalance the underlying.

Just chipping in (even though question wasn't directed at me), agree with ghchua. These smart-beta ETFs (or factor ETFs etc) have a few issues for retail investors:

1) while these factors have proven long-run outperformance, they can and will go through prolonged periods of underperformance (e.g. value has been underperforming for many years now). Even for a plain vanilla index ETF, some retail investors already find it hard to stick to the plan. Adding another layer of uncertainty makes it psychologically even tougher I think, unless you have a strong appreciation for how the factors work.


2) as ghchua pointed out, these vehicles have higher expense ratios, and it's all about trading off the extra cost against the possible factor premium that you receive. Take value as a factor for example, it has provided a historical premium of ~4%. But the premium refers to a long-short implementation (long cheap, short expensive), and most factor funds are long-only. So a ballpark premium that you will get is probably ~1-2%. This premium can only be harvested over the long run, and can even become negative in the short run. Further, the 4% premium was based on historical studies, and with the advent of so many smart beta products (it's not that hard to design such a product), it will not surprise me to see future premium be lower than the historical number. So you pay tens of basis points extra to harvest a ~1% premium over the long run, with periodic underperformance. Maybe it's worth it to some, but it's not a compelling idea for me

Side-note, Bogle himself hated such products (although Vanguard has pretty much ignored him and went ahead with their own smart beta products, I guess money is too important at the end of the day).

Hi ghchua and Corgitator, thank you for sharing your take on this.
I agree that increased fees is a concern for these so called smart beta ETFs, and it is probably not for people prone to chasing after the sexiest strategy at the moment.

But for investors who want to adhere to the strategy of a "defensive investor" described in the Intelligent Investor, perhaps we can regard the additional fees as a price to pay for the convenience of wide diversification into a portfolio of companies that are not overpriced based on metrics like PB, PE, etc. In other words, paying the fees not in the hope of outperformance, but rather as a safety feature to guard against any possible overvaluation among the companies with the biggest market cap (and thus over-represented in a market cap weighted index fund).

Sorry if I am deviating from the original thread intention of discussing about the general retail investor, but specuvestor's discussion about the possible problems in indexing struck a chord in me. I currently do active investing, but I can't rule out becoming more passive in future.
As many people have mentioned, many retail investors do not outperform the market, and I cannot rule out that I may be one of them.

weijian, thank you for the book recommendation. Will check it out.

I'm sure everyone has their own theory on value investing which is shaped by his/her own individual circumstances, but the challenge is how many of us actually track our investment TSR in a professional and precise way? For e.g. if someone ask you "How's has your value investment performance been for the past xx years?", are you able to answer with precision and confidence?

I understand that the few ways of measuring performance as follows:
1) Direct, basically an absolute % return
2a) Relative, out/under performance against chosen benchmark
2b) Relative, risk adjusted metrics sharpe, treynor, omega etc.

My quick polling around is that 80% probably don't even have a way to measure the simplest Direct performance, they seem to be just tracking daily/monthly size of their investment assets without even basic accounting for money in & out of the system, i.e. they can't even answer accurately what is their annual TSR over xx years.

Most of the remaining 20% usually have some general quick method such as XIRR function to estimate their performance, which is certainly better and offers a better view of performance, but still not really accurate. I myself only know of only 2 retail investors who are able to track performance precisely and from that basis calculate their performance in 2a/2b dimensions.

I feel this is something that severely clouds the judgement of retail investors. You can't really know whether your theory works unless you are able to demonstrate performance measurement of your portfolio over a sufficient period of time precisely. There is too much talk among retail investors over investment theories / stock evaluation and very little discussion on performance measurement. VB has some here and there, but then it's once again mostly dwarfed by the former.

The nature of performance management is of course more clouded and philosophical as one moves to higher realms of measurement, but no matter what I really think measuring one's investment portfolio in absolute TSR annually is the bare minimum competency anyone dabbling in the markets should have. Otherwise it's like running a business without any financial numbers and just relying on no. of customers or products sold to determine business viability.

(17-06-2020, 10:00 AM)mobo Wrote: I'm sure everyone has their own theory on value investing which is shaped by his/her own individual circumstances, but the challenge is how many of us actually track our investment TSR in a professional and precise way? For e.g. if someone ask you "How's has your value investment performance been for the past xx years?", are you able to answer with precision and confidence?

I understand that the few ways of measuring performance as follows:
1) Direct, basically an absolute % return
2a) Relative, out/under performance against chosen benchmark
2b) Relative, risk adjusted metrics sharpe, treynor, omega etc.

My quick polling around is that 80% probably don't even have a way to measure the simplest Direct performance, they seem to be just tracking daily/monthly size of their investment assets without even basic accounting for money in & out of the system, i.e. they can't even answer accurately what is their annual TSR over xx years.

Most of the remaining 20% usually have some general quick method such as XIRR function to estimate their performance, which is certainly better and offers a better view of performance, but still not really accurate. I myself only know of only 2 retail investors who are able to track performance precisely and from that basis calculate their performance in 2a/2b dimensions.

I feel this is something that severely clouds the judgement of retail investors. You can't really know whether your theory works unless you are able to demonstrate performance measurement of your portfolio over a sufficient period of time precisely. There is too much talk among retail investors over investment theories / stock evaluation and very little discussion on performance measurement. VB has some here and there, but then it's once again mostly dwarfed by the former.

The nature of performance management is of course more clouded and philosophical as one moves to higher realms of measurement, but no matter what I really think measuring one's investment portfolio in absolute TSR annually is the bare minimum competency anyone dabbling in the markets should have. Otherwise it's like running a business without any financial numbers and just relying on no. of customers or products sold to determine business viability.

Agree with you. Performance analytics is sorely lacking in most brokerages. For retail investors, I often suggest that they choose a broker with a robust performance analytics service (there's a few, but I won't name them because I don't want to be seen as advertising). It goes a long way in terms of assessing your performance as an investor.

(17-06-2020, 10:50 AM)Corgitator Wrote: Agree with you. Performance analytics is sorely lacking in most brokerages. For retail investors, I often suggest that they choose a broker with a robust performance analytics service (there's a few, but I won't name them because I don't want to be seen as advertising). It goes a long way in terms of assessing your performance as an investor.

Interactive Brokers (what I currently use) generates a pretty comprehensive performance analysis report. Where you can compare with benchmarks (including custom benchmarks). It's a 50+ page PDF that breaks down performance since inception (net of fees), including foreign exchange gains/loss, option gains/loss, realized gains/losses, accounting for impact of cash drag etc. https://www.interactivebrokers.com/en/index.php?f=5078

It uses a Modified Dietz Formula which approximates the true Internal Rate of return (IRR): https://www.interactivebrokers.com/image..._paper.pdf


Example portfolio (funny that even in an example portfolio, they had to show an investor under-performing a benchmark):

Not an ad; purely information sharing. (Moderator, please censor if violate any rules here)

Hi gzbkel

Most of my posts will indicate that I am skeptical on simple ratios or quant that doesn't go into details. Track record of Quants have not been spectacular over a cycle except Renaissance Tech's Medallion Fund which is like a black box. So I guess I better declare my bias first

I agree with most of the points VBs raised here but just to add some observations:

1) Averaging down on a broad index makes sense cause it is unlikely for an index, whether Thailand in GFC, Iceland or Greece in GFC or even South American countries' real economy to even halve. However the caveat is the FX adjustments for foreign investors. It makes sense for local investors. To a certain extent if the index goes to zero, maybe investments is not the biggest issue for the locals. It is somewhat similar to rolling a dice on 3, and then 2 and then 1. You might roll 1 next 10 rolls but it is not going to go down anymore, though it will cost you opportunity cost which I will elaborate in 3) .

However when you average down on a style or sector or theme that might not be the case. Nobody expected oil futures to go negative and a lot of ETF uses derivatives. Baltic Dry Index crashed 90% spectacularly in about 6 months. Those bullish on Tech during dot com might not have survived to ride on the past 5 years' bull run, even when they are right long term. Probably Singapore's best Quant Fund Quantedge lost 29% in 2018. If you had been buying a value styled fund, it would have been quite painful past 10 years, unless you are disciplined enough not to follow the Joneses

2) Fees Fees Fess. Yes you can imagine you are paying fees for diversification but so is ETF. So you got to find a good fund manager to be better than market but problem is I don't understand most quant's methodology or sometime like Madoff, they project that mystic about them. On the other hand in theory it is also not impossible to beat the market just using ETF and Asset Allocation, but I have yet to see one personally. The human Greed & Fear is very powerful, all the more when one does not have comfort on the underlying dynamics. That makes all the difference between knowing a company and knowing from a friend's recommendation.

3) I don't think Buffett actually uses a performance evaluator per se etc. Actually value investors are very upside absolute return driven. Hence the mantra "Don't lose money". When he compares to S&P in the Annual Report it is to compare against an Opportunity Cost, in this case you could have bought into a basket of S&P stoocks, or nowadays ETFs. But different people have different opportunity cost. It could be deposit rate, FD rate, WACC, lending rate if you are leveraging, IRR of a business venture etc. All depends on what would you have done instead. That's why Required Rate of Return in cash flow calculation is so fuzzy. To a certain extent it also explains when opportunity costs diminishes during ZIRP nowadays, asset prices goes up despite a recession. Unless you have a deflation like Japan.

In general my guess is to use a 7-10% threhold that has more than enough buffer. Seems to be what Chinese government been using as well to double GDP every 10 years.