Concentration is better than diversification

dub_nerd

Registered User
Messages
1,985
Here's a view which coincides with what I've been thinking for some time (and therefore may be an example of extreme confirmation bias on my part ;) ). Jim Rogers, co-founder with George Soros on the Quantum Fund says: "diversification is just something that brokers came up with, so they don't get sued. If you want to get rich ... you have to concentrate and focus."

https://news.thestreet.com/independ...ogers-on-why-not-to-diversify-portfolios.html

In this view, it's not all about length of time in the market. That'll just get you plodding returns (like the 2% my pension returned in the last year). My big gripe, though, is that diversification doesn't actually protect you. Yes, it'll save you from company- and industry-specific risk, but as long as you steer clear of stupid stuff the biggest risk by far is market risk which you can't really protect against.

So looking out for value buys and concentrating on a very small number of stocks seems to me to be reasonably sensible. So far I've managed to make a return of about 20% in six months while testing this strategy, though obviously that is much too short a time period to draw any inferences.
 
Here's a view which coincides with what I've been thinking for some time (and therefore may be an example of extreme confirmation bias on my part ;) ). Jim Rogers, co-founder with George Soros on the Quantum Fund says: "diversification is just something that brokers came up with, so they don't get sued. If you want to get rich ... you have to concentrate and focus."

https://news.thestreet.com/independ...ogers-on-why-not-to-diversify-portfolios.html

In this view, it's not all about length of time in the market. That'll just get you plodding returns (like the 2% my pension returned in the last year). My big gripe, though, is that diversification doesn't actually protect you. Yes, it'll save you from company- and industry-specific risk, but as long as you steer clear of stupid stuff the biggest risk by far is market risk which you can't really protect against.

So looking out for value buys and concentrating on a very small number of stocks seems to me to be reasonably sensible. So far I've managed to make a return of about 20% in six months while testing this strategy, though obviously that is much too short a time period to draw any inferences.

You can be a passive investor or a stock picker. They are two completely different things.

Of course diversification will not help you get excellent returns, thats like saying, this golf club doesnt make my rally car go any faster. For a passive investor diversification is a free lunch, it protects you against certain risks, free of charge, that's a marvellous thing. It can protect you against business risk, against industry risk, against currency risk, against geographical risk. Its true that it cannot protect you against systemic risk, but thats like saying, this golf club was useless when the golf course flooded.

But you didn't want to play golf, you wanted to go rallying.

Absolutely diversification has little value for the stock picker. With stock picking why should you put money in the second best investment opportunity? However it is a simple matter of arithmetic that the average euro invested by a stock picker will make a market return and less after costs. Of course that average hides many above average returns, Jim Rodgers for example, and many below average returns, you for example ?

More than golf or rallying, stock picking is like playing chess, anyone can learn how its done, lots of study and time can make you better at it, not many people become world champions.
 
So looking out for value buys and concentrating on a very small number of stocks seems to me to be reasonably sensible.

No, it's not sensible at all.

It might be for Jim Rogers or Warren Buffett, but it's crazy for 99.99% of people.

So far I've managed to make a return of about 20% in six months while testing this strategy, though obviously that is much too short a time period to draw any inferences.

The fact that you would even refer to 6 months' returns, suggests to me that you should not be doing anything other than investing in a broad, diversified, portfolio of shares.

Brendan
 
Here's my take on how to "get rich":

- Take a big concentrated binary bet in a sector you know like the back of your hand (e.g. set up a business), or

- Optimise your income and your assets from an asset allocation and tax perspective and then invest in a diversified manner.

Think about the two great tax breaks, PPR Relief and pensions. The power to compound wealth tax free in a pension is huge. And it's easy to forget that 7% a year doubles your money after 10 years. You don't need to achieve 20% returns (and to do so, the risks are off the scale).
 
Can't be certain, but I'm not sensing a whole lot of agreement with my position. :D

Just added another batch of shares to my highly non-diversified portfolio. :eek:
Will let you know how it goes next year. ;)
 
You will have to wait around 20 years to have any meaningful evidence that any out-performance or under-performance is due to anything other than the randomness of returns.

Brendan

I dont agree with this at all. The whole point about picking a single share is that it avoids the randomness of returns.

If I think share X is undervalued and buy into it with a 6 months time horizon, I am entitled to regard the profit or loss I make beyond the market return as down to my skill and courage. Luck of course plays a part but that is true in any area of human endeavour.

One success or failure of course does not prove that there will be a consistent pattern into the future.
 
I just read the title on this new thread.

Concentration is better than diversification. Well given the season thats in it, the obvious answer is, "Oh No it isn't"

If fact I think the point is that they are two completely different things, diversification is an excellent for a passive investor. For a successful stock picker diversification will put a drag on returns. Wether, you me or anyone else will be a successful stocker into the future I have no idea.
 
I dont agree with this at all. The whole point about picking a single share is that it avoids the randomness of returns.

If I think share X is undervalued and buy into it with a 6 months time horizon, I am entitled to regard the profit or loss I make beyond the market return as down to my skill and courage. Luck of course plays a part but that is true in any area of human endeavour.

One success or failure of course does not prove that there will be a consistent pattern into the future.

What it doesn't avoid is potential wealth devastation. Plenty of excellent companies have been hammered because of one-off issues (e.g. Volkswagen with emissions, Tesco with accounting). Apple's a great company, but who's to say that smartphones won't start exploding or be linked to dementia?

People in other countries (e.g. the UK/US) where there is a greater history of wealth preservation and wealth creation have it copped. Aim to compound away at 5-7% over time and you'll be laughing.
 
You will have to wait around 20 years to have any meaningful evidence that any out-performance or under-performance is due to anything other than the randomness of returns.

Brendan
I may be a naive and inexperienced investor but I do know my maths. :)
You'd have to define "meaningful" and then pick a metric but, for example, the chances of outperforming the market every year for twenty years by random chance are a million to one. We can use combinatorics to assign a probability to any given number of years of outperformance, e.g. a one in four chance of outperforming for 12 or more years out of 20. That's just assigning a binary chance of being higher or lower than the average market each year. If you assign weights to the degree of outperformance you should be able to make a determination in much less than 20 years. Obviously it's always impossible to rule out pure dumb luck -- that's the nature of probability.
 
Six months? That's almost like saying you're the better team when you're one nil up after 2 minutes. The game has only just begun.
If you're one-nil up after two minutes it doesn't make you the better team, but it certainly increases the probability that you are. What level of evidence would you consider significant?
 
"diversification is just something that brokers came up with, so they don't get sued. If you want to get rich ... you have to concentrate and focus."

He's right. Diversification reduces risk exposure. No one can predict the top performing region or sector year on year. Diversification merely reduces the wide volatility that you are exposed to by having a concentrated portfolio.

If you had two funds, one returned 10% each year for 10 years. The other returned 20% for 7 years and lost 20% for another 3 (sequence doesn't matter), which would you go for?

The 10% option returns over 40% of the 20% option. Concentrated stock picking is liable to give you the same result. You can get the really exciting highs but you are very open to the devastating lows too (like a drug!!).

Then we get to have you got the skill to pick the correct stock? Active fund managers have highly skilled people and millions in resources. They talk to CEO's, they talk to their competitors, the do tons of research before buying stock. And even still, they get things wrong a lot of the time. You are competing against these professionals. Are you going to put the same time and effort into analysing a company before buying? Where are you going to start?


Steven
www.bluewaterfp.ie
 
If you're one-nil up after two minutes it doesn't make you the better team, but it certainly increases the probability that you are. What level of evidence would you consider significant?
A full season. 90 minutes * 38 games.

In investing that would be 20 years at least.
 
If you're one-nil up after two minutes it doesn't make you the better team, but it certainly increases the probability that you are. What level of evidence would you consider significant?

A full season. 90 minutes * 38 games.

In investing that would be 20 years at least.

Ok, so not trying to be smart or anything, but now let's get down to where you plucked 20 years from (same as Brendan previously). Is it because it sounds like a long time? Let's try to be a bit more scientific. If someone -- using whatever strategy of their choice -- has a random chance of being ahead of or behind the market at the end of each year, then the chances are a million to one against them beating the market every year for 20 years in a row (a million being the twentieth power of two). I don't think we need anything like that level of certainty to demonstrate that their strategy has something going for it. I'm not even suggesting that this binary end-of-year measurement is the right way to do it, but we have to have some metric, otherwise it's just completely arbitrary.
 
Take the US market over the course of its history (I think it's 1871 to present day).

Taking a one day time horizon, and all observations over the period, investors have been up 52% of the time. Not far off a coin toss.

However, taking any 20 year time horizon, an investor has NEVER been down.
 
Concentrated stock picking is liable to give you the same result. You can get the really exciting highs but you are very open to the devastating lows too (like a drug!!).
Good point, Steven. Maybe I'm just trying to inject a bit of stimulation. I do have to admit the thought of slapping my cash in someone else's fund makes my eyes glaze over a bit. I want to have a bit of fun with this.

Then we get to have you got the skill to pick the correct stock? Active fund managers have highly skilled people and millions in resources. They talk to CEO's, they talk to their competitors, the do tons of research before buying stock. And even still, they get things wrong a lot of the time. You are competing against these professionals. Are you going to put the same time and effort into analysing a company before buying? Where are you going to start?
Also good questions. Part of my answer is that the fund manager has to stay invested all the time. To quote one of them: "there's no other game in town". I, on the other hand, don't have to. My needs are few. I can cherrypick the best value buys or I can stay away altogether. Ok, I still have to figure out what they are, but with only needing to trade a handful of times a year I can buy the significant dips (along with learning a bit of analysis along the way). And let's not forget that most active fund managers don't outperform the market either.
 
Ok, so not trying to be smart or anything, but now let's get down to where you plucked 20 years from (same as Brendan previously). Is it because it sounds like a long time? Let's try to be a bit more scientific. If someone -- using whatever strategy of their choice -- has a random chance of being ahead of or behind the market at the end of each year, then the chances are a million to one against them beating the market every year for 20 years in a row (a million being the twentieth power of two). I don't think we need anything like that level of certainty to demonstrate that their strategy has something going for it. I'm not even suggesting that this binary end-of-year measurement is the right way to do it, but we have to have some metric, otherwise it's just completely arbitrary.
20 years is absolutely arbitrary. As is 6 months. There's nothing scientific about it.

You mention 'beating the market every year'. Why are you using an arbitrary time frame like a year? It just happens to be the time it takes the earth to orbit the sun. What has it got to do with investing?

I suppose the 'best' time frame is the amount of time it takes to reach your investing goal. If you're able to cash out and retire with market beating performance then I'd say you're strategy has worked. Until then it's still under investigation.

Take the US market over the course of its history (I think it's 1871 to present day).

Taking a one day time horizon, and all observations over the period, investors have been up 52% of the time. Not far off a coin toss.

However, taking any 20 year time horizon, an investor has NEVER been down.

This is an excellent reason to use a 20 year time frame.
 
The title of this thread is concentration vs diversification.

Diversification is part of accepting the idea of the efficient market.

There are two different reasons why you might move away from the efficient market idea.

1. To an alternative market view, which involves a rule or scheme to underpin investment choices. Value investing, growth investing, dogs of the dow, sell in May, are all examples. While they may move away from full diversification none of these involve concentration as desirable in itself. All of these involve the idea that there is a better way to value shares on a systematic basis than the efficient market theory. A long term test makes some sense toward such an approach.

2. To the idea that on a share by share basis you can sometimes spot errors in the market price. There is no rule or scheme involved here. Just a view that share X is mispriced by the market. One share may be mispriced for one reason another share mispriced for a different reason. You might be right on one occasion and wrong on the next occasion, there is no valid test for such an approach. Concentration is essential to this approach.

The question which is better, diversification or concentration, is based on a false premise. Each has a different function.

If you want the markets to work for you then you should diversify. If you think you can identify market errors, then you must concentrate.
 
Back
Top