Publications
Partners
Concentration risk is one of the biggest risks an active investor faces. Sometimes it’s easy to manage, but at other times this is not the case.
I stress this is mostly an active investor problem; a passive investor holding an exchange-traded fund (ETF)
automatically has a broad spectrum of shares across different industries, geographies and to a lesser degree asset classes (as property is likely to be included).
Locally, however, passive investors investing in the Top40 do have some serious concentration risk, with Naspers* making up over 20% of the index, while Richemont** makes up another 10%.
This is great when these stocks are running higher, but when they turn things could get nasty. This is why I prefer an equal-weighted ETF locally.
The primary local ETF I own is the CSEW40, which has the exact same shares as the Top40, but with a weighting of 2.5% each – significantly downweighting Naspers and Richemont.
It does mean that this ETF has underperformed a vanilla Top40 ETF over the past few years – but I am comfortable with that. The point is, we love the concentration when a stock is running hard. We feel smart and of course are making money.
But as it runs, it also increases its weighting within our own active portfolio, which means we are at higher risk due to one single share – something that runs against the grain of portfolio construction.
An ideal portfolio should have at least 10 to 12 shares, and this is itself very concentrated, but I am assuming that at least half of the money invested into the market is already in ETFs, markedly reducing overall concentration risk.
If you only hold individual stocks and no ETFs, then 20 stocks are the minimum for a diverse portfolio. But the problem remains: what do we do when one of them is an absolute winner? This happened to me when I first bought Capitec.
It was around 2 000c a share and I put it in at 10% of my active portfolio. Capitec was a star performer, so much so that it was soon a 50% weighting of my individual stocks! One stock as half a portfolio is insane risk.
So how do we manage that risk? I sold some Capitec and diverted all new money and cash from the sales into other stocks.
But given the speed at which Capitec was rising, I was struggling to manage the weighting and the net result is that, aside from my initial purchases at 2 000c and 4 000c, and then some more at around R205 when African Bank failed, I have spent much of the past decade selling Capitec.
I’ve been selling my best-performing stock, yet it is still one of my top holdings percentage-wise. The concern is that selling winners is not a great idea, but we do need to protect our portfolio from that concentration risk.
What if Capitec had crashed while it constituted 50% of my portfolio? If it lost half its value, my portfolio would have fallen 25%. So, I have rules in place. I start with a 10% weighting and I am happy to allow the stock to get to a 30% weighting; above that I start selling down the position.
If the holding is below 20%, I will buy more if it is considered cheap by my metrics and does not go above a 20% weighting. When the stock hits 20%, I stop buying, and over 30% I start selling.
The truth is that I hate selling winners, but I need to manage the concentration risk.
*finweek is a publication of Media24, a subsidiary of Naspers.
**The writer owns shares in Richemont, Capitec and CSEW40.
I stress this is mostly an active investor problem; a passive investor holding an exchange-traded fund (ETF)
automatically has a broad spectrum of shares across different industries, geographies and to a lesser degree asset classes (as property is likely to be included).
Locally, however, passive investors investing in the Top40 do have some serious concentration risk, with Naspers* making up over 20% of the index, while Richemont** makes up another 10%.
This is great when these stocks are running higher, but when they turn things could get nasty. This is why I prefer an equal-weighted ETF locally.
The primary local ETF I own is the CSEW40, which has the exact same shares as the Top40, but with a weighting of 2.5% each – significantly downweighting Naspers and Richemont.
It does mean that this ETF has underperformed a vanilla Top40 ETF over the past few years – but I am comfortable with that. The point is, we love the concentration when a stock is running hard. We feel smart and of course are making money.
But as it runs, it also increases its weighting within our own active portfolio, which means we are at higher risk due to one single share – something that runs against the grain of portfolio construction.
An ideal portfolio should have at least 10 to 12 shares, and this is itself very concentrated, but I am assuming that at least half of the money invested into the market is already in ETFs, markedly reducing overall concentration risk.
If you only hold individual stocks and no ETFs, then 20 stocks are the minimum for a diverse portfolio. But the problem remains: what do we do when one of them is an absolute winner? This happened to me when I first bought Capitec.
It was around 2 000c a share and I put it in at 10% of my active portfolio. Capitec was a star performer, so much so that it was soon a 50% weighting of my individual stocks! One stock as half a portfolio is insane risk.
So how do we manage that risk? I sold some Capitec and diverted all new money and cash from the sales into other stocks.
But given the speed at which Capitec was rising, I was struggling to manage the weighting and the net result is that, aside from my initial purchases at 2 000c and 4 000c, and then some more at around R205 when African Bank failed, I have spent much of the past decade selling Capitec.
I’ve been selling my best-performing stock, yet it is still one of my top holdings percentage-wise. The concern is that selling winners is not a great idea, but we do need to protect our portfolio from that concentration risk.
What if Capitec had crashed while it constituted 50% of my portfolio? If it lost half its value, my portfolio would have fallen 25%. So, I have rules in place. I start with a 10% weighting and I am happy to allow the stock to get to a 30% weighting; above that I start selling down the position.
If the holding is below 20%, I will buy more if it is considered cheap by my metrics and does not go above a 20% weighting. When the stock hits 20%, I stop buying, and over 30% I start selling.
The truth is that I hate selling winners, but I need to manage the concentration risk.
*finweek is a publication of Media24, a subsidiary of Naspers.
**The writer owns shares in Richemont, Capitec and CSEW40.
