Rebalancing, really?

[EssDeeAitch]

I am aware that there are proponents of rebalancing a portfolio (restated by a Patrick Connolly of Chase De Vere in this weeks IC on the Portfolio Clinic column), the principle being to "take profits from investments that have done well and reinvest the proceeds in ones that have done badly".

This to me is counter-intuitive as surely one should keep backing the risers and sell the fallers? So the question is, "is this a good investment strategy"?

[Bubblesofearth]

No it's not

[ADrunkenMarcus]

When I'm gardening, I always water and nourish the weeds and cut back or kill the healthy, growing plants.

Best wishes

Mark.

[SalvorHardin]

It's a poor strategy to put it mildly. Generally the opposite is better. Cut your losses and let your profits run.

The idea of selling winners and buying losers is rooted firmly in loss aversion and prospect theory, two of the major ideas of behavioural economics. Humans prioritise avoiding losses to the point where the psychological benefit from making a gain of 1 unit will not offset the psychological pain of losing 1 unit (generally gains need to be at least twice the losses for the psychological effects to cancel each other). There's also a bit of "reversion to the mean", where it is expected that the good performing share will fall back to the market average whilst the badly performing share will rise back to the market average.

Selling at a loss means admitting that you got it wrong. Many people would prefer to jump through flaming hoops than to admit that they made a mistake, even if they only admit this to themselves (not just about investment matters, just look at the entrenched views in politics, sports, etc.). This is why so many investors get stuck in loss-making positions, clinging onto shares in companies whose business is falling apart. They refuse to sell until the investment has recovered to the price they paid. Even if they discover a better investment (as they see it) they refuse to sell the loser and reinvest the proceeds in the new investment.

Of course there are exceptions; sometimes an investment's market price is well below its intrinsic value (or the value which you determined it was worth buying) simply because of market movements or that investors' opinion is wrong. See Benjamin Graham's story about Mr. Market, if you've never heard of it I strongly recommend it - see the link at the end of this post). But generally clinging onto losing positions for many years is a poor strategy. There's a psychological benefit to be gained from dumping a long-term loser, as many long standing investors will know all too well. Just avoid looking at the price of a share after you've sold it

A phrase you'll encounter regarding losses on shares is that you've only made a loss if you sell (i.e. it's only a paper loss). People use this to justify holding on whereas in reality they have lost money; they just haven't crystallised the loss. But if these shares were pledged as security for a loan there's no way that the lender would use the original purchase price instead of the market price.

Part of the strategy is rooted in trimming your gains and increasing diversification to avoid a great performing share from dominating your holdings to the extent that your finances become dependent upon it. Now there's nothing wrong it doing a bit of trimming, particularly if you've got so much in the one company that it starts to affect your sleep (this happened to me in 2008, though it was three medium sized oil explorers not just the one share, and I ended up trimming in order to sleep better).

https://en.wikipedia.org/wiki/Mr._Market

https://en.wikipedia.org/wiki/Loss_aversion

https://en.wikipedia.org/wiki/Prospect_theory

[Urbandreamer]

This to me is counter-intuitive as surely one should keep backing the risers and sell the fallers? So the question is, "is this a good investment strategy"?

Well it depends upon what you are trying to achieve. If you are seeking to grow your capital, then obviously the stratergy has its problems. If however you are attempting to preserve your wealth then there may be more merrit.

Consider a situation where you are 100% equity invested in 5 shares (obviously you should be more diversified). One has done very well and is now 50% of your wealth (you definatly should be more diversified). What are the odds that it will continue its rise and what the odds that it will stumble and fall.

Consider a share that I got right.
https://www.londonstockexchange.com/exc ... XSTMM.html
I first bought in 2011. The gains meant that it became a larger and larger part of my portfolio.
I trimmed in Feb and agian in June this year.

I didn't know and couldn't predict the August fall and Trump tariffs, but a drop in the share price of that one share would have had a noticable effect upon my portfolio.

So, do you think that there is a right or wrong answer to your question irrispective of why someone has a portfolio of investments?

[tjh290633]

I am aware that there are proponents of rebalancing a portfolio (restated by a Patrick Connolly of Chase De Vere in this weeks IC on the Portfolio Clinic column), the principle being to "take profits from investments that have done well and reinvest the proceeds in ones that have done badly".

This to me is counter-intuitive as surely one should keep backing the risers and sell the fallers? So the question is, "is this a good investment strategy"?

There is another way of looking at this. If you are going to rebalance it implies that one or more shares have become very overweight and others are underweight. That does not necessarily mean that the underweight shares have done badly. The market always has a wide spread of movements, sometimes what does well one year does badly next year and vice versa.

To those looking for income, the high flying shares will usually have lower yields than those lagging behind. It therefore makes sense to trim back the overweight shares and reinvest in your higher yielding shares. It may also make sense to sell completely a share whose yield has fallen to a low level, say below half the market yield, and buy a replacement share with a yield above the market average. In this way you can ratchet up the income well above the natural growth rate.

If the objective is purely capital growth, then different considerations will apply.

TJH

[colin]

This to me is counter-intuitive as surely one should keep backing the risers and sell the fallers? So the question is, "is this a good investment strategy"?

Not if you invest in individual company shares, absolutely yes if you hold cap weighted index trackers. Surprised that so called investors can't understand this.

[EssDeeAitch]

Thanks for all the responses, they are very helpful. To me, the major issue is the STRATEGIC imperative of rebalancing (routinely selling or top slicing the best performers to increase allocations to lower performers) which I find wrongheaded and the TACTICAL selling of a stock once either an objective has been reached or fear approaches.

My reading of the replies would seem to share this opinion.

[SalvorHardin]

"Selling your winners and holding your losers is like cutting the flowers and watering the weeds." - Peter Lynch

"Suggesting that an investor should sell portions of their best performing shareholdings simply because they have become such a major part of their portfolio is akin to suggesting that Barcelona should sell Lionel Messi because he has become so important to the team" - "adapted from Warren Buffett's 1996 Berkshire Hathaway shareholders' letter by changing the sportsman (Michael Jordan) to make it more relevant for a British audience

[Dod101]

I am well aware of the quotes but the other side of this argument is what has happened to BAT £50 to around £27. Trimming BAT (and we had plenty of time to do so because it was well over £40 for a long while) could have saved a lot of anguish. Of course if you are a Buffett you just ignore the market gyrations because of course, he can afford to.

I have a number of relatively small holdings which could do with a top up and rebalancing can of course make sense, because these smaller holdings may not be my best shares but they are not losers (and certainly not necessarily losers) so all this talk of comparing rebalancing as like cutting down the flowering shrubs and cultivating the weeds is colourful and fun but need not be the case. IN any case even the best shrub needs to be trimmed occasionally. I think there is quite a lot to be said for rebalancing, although I have to admit it does sometimes feel counter intuitive.

'Locking in' or taking profits and recycling them is another way to look at it. Feels better thinking of it that way, to me anyway.

Dod

[EssDeeAitch]

I am well aware of the quotes but the other side of this argument is what has happened to BAT £50 to around £27. Trimming BAT (and we had plenty of time to do so because it was well over £40 for a long while) could have saved a lot of anguish. Of course if you are a Buffett you just ignore the market gyrations because of course, he can afford to.

I have a number of relatively small holdings which could do with a top up and rebalancing can of course make sense, because these smaller holdings may not be my best shares but they are not losers (and certainly not necessarily losers) so all this talk of comparing rebalancing as like cutting down the flowering shrubs and cultivating the weeds is colourful and fun but need not be the case. IN any case even the best shrub needs to be trimmed occasionally. I think there is quite a lot to be said for rebalancing, although I have to admit it does sometimes feel counter intuitive.

'Locking in' or taking profits and recycling them is another way to look at it. Feels better thinking of it that way, to me anyway.

Dod

Hi Dod, agree with you completely in that tactical selling and reallocation (to existing or new investments) must be a good thing as opposed to top slicing or selling out simply to rebalance.

[argoal]

I am aware that there are proponents of rebalancing a portfolio

This to me is counter-intuitive as surely one should keep backing the risers and sell the fallers? So the question is, "is this a good investment strategy"?

This discussion reminds me of one the the long unsolved problems in physics - The Theory of Everything.

The nub of the problem is that general relativity describes perfectly the behaviour of large objects and quantum field theory explains the behaviour of sub-atomic particles.

Both theories are incredibly accurate when applied to their respective domains. However the frameworks they are built on are incompatible so they both can't be right. The Theory of Everything is an attempt to construct one framework that works accurately in both domains.

Bear with me, but this domain problem, it seems to me, also applies to the efficacy of rebalancing within a portfolio.

There is copious literature confirming that at a portfolio level rebalancing between asset classes provides higher a risk adjusted return than a non balanced portfolio (note risk-adjusted). It has also been shown to support a higher withdrawal rate in deaccumulation and, as Terry points out, a higher income in a yield based portfolio.

However, at an individual stock level, the momentum factor has been shown to be persistent and real. That is, in effect, running winners and shorting losing stocks. Hence there is the paradox that rebalancing is both a good and bad strategy at once.

The reason may be that individual stock values are less likely to mean revert than asset classes. Winning stocks tend to keep winning while stocks that have lost 50% of value are just as likelt to keep falling. asset classes, on the other hand tend to have their 'day in the sun' periodically. The cyclical nature of asset performance may give rise to the difference between the portfolio and individual benefits of rebalancing.

So it is possible that Patrick Connolly is both right and also wrong depending on the domain being considered.

For disclosure - I am neither a physicist nor an investment professional. So the above may well be nonsense.

[colin]

For disclosure - I am neither a physicist nor an investment professional. So the above may well be nonsense.
Top

No it's not nonsense, well to me at least. Einsteins theories rely on nothing traveling faster in a vacuum than Maxwells fixed speed of electromagnetic radiation, whereas Quantum theory allows for what Einstein called 'spooky action at a distance' whereby a change in the state of an electromagnetic particle is instantaneously accompanied by a change in the state of a particle possibly thousands of miles away to which it is 'Quantumly entangled'.
Both theoretical predictions have been proved true by experimental observation.

Those who invest in cap weighted index trackers have eliminated specific company risk and taken on market risk which is mostly determined by economic cycles, so rebalancing between non-correlated asset classes can reduce overall volatility, for those who invest in individual shares the smaller the number of companies in the portfolio the more company specific becomes the risk, and however large the portfolio rebalancing into shares with the poorest performing price can lead one into pouring money into 'falling knife' companies, in other words the worst businesses in a portfolio.

[argoal]

No it's not nonsense, well to me at least. Einsteins theories rely on nothing traveling faster in a vacuum than Maxwells fixed speed of electromagnetic radiation, whereas quantum theory allows for what Einstein calls 'spooky action at a distance' whereby a change in the state of an electromagnetic particle is instantaneously accompanied by a change in the state of a particle possibly thousands of miles away to which it is 'Quantumly entangled'.
Both theoretical predictions have been proved true by experimental observation.

Those who invest in cap weighted index trackers have eliminated specific company risk and take on market risk which is mostly determined by economic cycles, so rebalancing between non-correlated asset classes can reduce overall volatility, for those who invest in individual shares the smaller the number of companies in the portfolio the more company specific becomes the risk, and rebalancing can lead one into pouring money into 'falling knife' companies, in other words the worst investment companies in a portfolio.

Wot he said

[StepOne]

I am well aware of the quotes but the other side of this argument is what has happened to BAT £50 to around £27. Trimming BAT (and we had plenty of time to do so because it was well over £40 for a long while) could have saved a lot of anguish.

You make it sound easy. I trimmed BATS in 2009, when it hit £17. It's delivered over £13 in dividends since then, never mind the 160% capital gain, so not a great outcome that time.

StepOne

[Gengulphus]

I am aware that there are proponents of rebalancing a portfolio (restated by a Patrick Connolly of Chase De Vere in this weeks IC on the Portfolio Clinic column), the principle being to "take profits from investments that have done well and reinvest the proceeds in ones that have done badly".

Maybe that is the principle used by some proponents of rebalancing a portfolio (such as Patrick Connolly?) but it certainly isn't that used by all of them. For example, I do some rebalancing of my portfolio and I am a proponent of doing so in a rather weak sense - essentially, my belief is that if imbalances grow big enough, practically every investor will find themselves wanting to do it: with only very rare exceptions (if any), the issue isn't whether an investor will decide to rebalance, but how big an imbalance is needed to get them to the point of making that decision.

That's in principle, by the way: in practice, it might be extremely unlikely that the imbalance will grow big enough during their lifetime... But "extremely unlikely" isn't the same thing as "impossible" - for instance, some people found that the tech boom caused imbalances to grow extremely big in just a few years leading up to the first few months of 2000. The sensible ones rebalanced to at least some extent; the less sensible ones didn't and lost a very high fraction of their gains as the bubble burst - and that's true of even the basically sound companies: for instance, ARM Holdings floated at under 50p in the spring of 1998, peaked about two years later at around £10 in the spring of 2000; by about 2.5 years later in the autumn of 2002, it had fallen back to below 50p (all prices split-adjusted). So roughly speaking, those who just held throughout lost all the gains they made in the tech boom when it burst (and that's for what subsequently turned out to be one of the best UK tech shares - investors in plenty of others didn't just lose all their tech boom gains, but also their original investment as the companies went bust) while those who rebalanced by selling even 10% of their holding in early 2000 and reinvested the proceeds in 'old economy' shares like utilities, miners, banks, etc, took out a multiple of their original investment and then probably made some nice gains on the proceeds, as (not very surprisingly) 'old economy' shares were at a low ebb at the height of the tech boom.

Anyway, the rebalancing principle I use is definitely not "take profits from investments that have done well and reinvest the proceeds in ones that have done badly". Instead, it can be stated in a similarly summarised way as "take profits from holdings that have become uncomfortably big and reinvest the proceeds in holdings that one thinks will do well in future and are currently well under being uncomfortably big". For that last part, by the way, treat non-existent holdings (i.e. new shares) as though they were zero-size existing holdings and thus automatically well under being uncomfortably big.

Note that those two rebalancing principles can lead to very different answers. For instance, imagine my portfolio contains two holdings, of shares A and B. Share A was one that I was reasonably happy to invest in originally and I put 1% of my portfolio value into it; share B was one I was less happy with originally, but that I reckoned was worth a flutter, and I put 0.1% of my portfolio value into it. Both shares have in fact done very well, so that my holding of share A is now 10% of my total portfolio value and my holding of share B 1% of my portfolio value - and their future prospects look very good to me, so I'm now very happy to be invested in both. Now look at what the two rebalancing principles say:

* The "take profits from investments that have done well and reinvest the proceeds in ones that have done badly" principle says shares A and B have both done well, not badly, so I should take profits from both and reinvest the proceeds in other shares that have done badly.

* The "take profits from holdings that have become uncomfortably big and reinvest the proceeds in holdings that one thinks will do well in future and are currently well under being uncomfortably big" principle requires knowledge of how big a holding needs to be before I find it uncomfortably big. For the sake of this example, assume that I consider holdings worth more than 5% of the portfolio value to be uncomfortably big (this isn't actually correct for me, but it's a round number in the right general area and it's considerably simpler than the correct answer). So the holding of share A is uncomfortably big for me and my principle says that I should take profits from it, while the holding of share B is not just not uncomfortably big, but well short of being uncomfortably big. Since I'm also very happy about its future prospects, the principle says not just that I shouldn't take profits from it, but also that I should reinvest some of the proceeds from the sale of share A into it. Though not all of them, since those proceeds would be at least 5% of the portfolio value and added to the existing holding, would make the holding worth 6% of the portfolio value and so uncomfortably big. A plausible use of the sales proceeds would be to put 2% of the portfolio value into share B, raising the holding to 3% of the portfolio value (which allows a reasonable amount of room for future growth before it becomes uncomfortably big), and find somewhere else for the remaining 3%+.

Note that although the fact that I use it obviously implies that I consider my rebalancing principle better than "take profits from investments that have done well and reinvest the proceeds in ones that have done badly", at least for me, I'm not posting here to argue whether it's better or worse. Just that it's different from the principle you stated, both as regards its statement and what it tells one to do in practice, and I at least (quite possibly others as well) simply don't feel I can sensibly discuss rebalancing on the basis of the principle stated in your OP.

Or in short, my personal answer to the thread title is "Yes, really rebalancing - but rebalancing as you've described it, no way!"

Gengulphus

[Dod101]

You make it sound easy. I trimmed BATS in 2009, when it hit £17. It's delivered over £13 in dividends since then, never mind the 160% capital gain, so not a great outcome that time.

None of this is easy and I did not trim BATS when I could have done at £50! I have held BATS since July 2001 when I bought them at £5.4475. I have been repaid in spades since then and for what it is worth am now getting well over 30% yield on my original purchase price.

Dod

[Gengulphus]

For disclosure - I am neither a physicist nor an investment professional. So the above may well be nonsense.
Top

No it's not nonsense, well to me at least. Einsteins theories rely on nothing traveling faster in a vacuum than Maxwells fixed speed of electromagnetic radiation, whereas Quantum theory allows for what Einstein called 'spooky action at a distance' whereby a change in the state of an electromagnetic particle is instantaneously accompanied by a change in the state of a particle possibly thousands of miles away to which it is 'Quantumly entangled'.
Both theoretical predictions have been proved true by experimental observation.

No, not quite, because no scientific theory is ever "proved true by experimental observation". What is the case is instead "Neither theoretical prediction has yet been proved false by experimental observation, despite really serious attempts." Which may seem pedantic, but it carries the key to how the contradiction between the two might eventually be resolved. In particular, the contradiction implies that in principle, there are experiments in which one theory says that X happens and the other says that Y happens (Y being incompatible with X). If the theories had both indeed been "proved true", then if and when the experiment is actually performed in practice, the resulting experimental observation would show both X and Y happening, which isn't possible because they're incompatible. But with the situation actually being that neither theory has yet been proved false, there are the possibilities that the experimental observation shows X happening (proving at long last that the second theory is false), or that it shows Y happening (proving at long last that the first theory is false), or that it shows Z happening (proving at long last that both theories are false).

There are other possible resolutions - for instance, it might be that it's actually impossible to perform any such experiment in practice. Though any theory that that is impossible again cannot be proved true - the most one might manage is "No way has ever yet been found to perform such an experiment in practice, despite really serious attempts"... The thinking concerned definitely becomes complex and contorted!

How (if at all) this applies to theorising about investment strategies I'll leave to those more skilled than me, other than to observe that good experiments on investment markets and strategies are very hard to perform in practice, given the years- and often decades-long periods needed to gather statistically significant amounts of data and the tendency for theories that produce useful predictions to carry the seeds of their own disproof if they become well-known...

Gengulphus

[hiriskpaul]

Those who invest in cap weighted index trackers have eliminated specific company risk and taken on market risk which is mostly determined by economic cycles, so rebalancing between non-correlated asset classes can reduce overall volatility, for those who invest in individual shares the smaller the number of companies in the portfolio the more company specific becomes the risk, and however large the portfolio rebalancing into shares with the poorest performing price can lead one into pouring money into 'falling knife' companies, in other words the worst businesses in a portfolio.

Agree 100% with the first part regarding rebalancing between non-correlated assets (or at least low correlation), but the second part is a little more nuanced. Mindlessly selling your winners to buy your losers is I agree a silly thing to do. Equally though, mindlessly selling your losers to buy into your winners can also lead to a bad outcome. Obvious example, selling your old fashioned "loser" non-technology stocks and ploughing into your "winning" technology stocks during the dot-com boom/bust.

There can be huge value in losers, so why mindlessly sell them? For example, anyone holding Apple shares between 2000 and 2003 would have lost around 80% of their money, but over the long term buying more of (or just holding on to) this particular "loser/falling knife" would have been a very smart move. I have found investing in losers/falling knives very profitable and much prefer it to investing in winners, although I mostly focus on the loser's debt securities.

For those running an active portfolio instead of following mantras, such as sell your losers, run your winners, take a regular look at your portfolio with a view to which stocks are worth holding/adding to at the current price and which are not or can be reduced, but factor in your single company/sector/market exposure. This factoring in is where rebalancing potentially comes in. Pay no attention at all to the price you paid either, or for that matter, the price it was at last week unless you are a momentum investor/trader, as these are worthless pieces of information.

[colin]

Obvious example,

aren't all examples obvious with hindsight?