Is ten Irish shares enough?

The need to diversify across asset types, and across equities comes to mind. Brendan you propose to 'index' track the top 10 Irish stocks. I prefer to advise diversification across the biggies in Europe, since Ireland's stocks are still only medium at best and represent a market 1% of Europe.

Given the higher stock select risk of the top 10 Irish pick would you now alter your view and suggest a basket of 20 to 50 Euro stocks, using an index?

By the way Brendan, like many other folk I've suffered too from these stories, particularly Elan. sure I know they'll recover and are good long term bets, but what about the inherent higher risk levels relative to say the Eurostoxx indices?

Re: elan/aib

Hi Mith

I have never suggested index tracking. I suggested buying the top 10 stocks in equal amounts. Tracking gives rise to overconcentration e.g. at one stage you would have had 25% of your money in Vodafone if you tracked the FTSE 100.

That aside, there are two issues: Is ten enough? and Ireland vs. Europe.

<!--EZCODE BOLD START--> Is ten enough?<!--EZCODE BOLD END-->
The way people are talking you would think that all ten stocks have gone into liquidation. Elan has lost about 80% of its value, so a portfolio containing Elan has lost 8% of its value. That's not nice, but it's not the end of the World. A portfolio containing 20 shares would still have lost 4% of its value.

What is confusing people is that both have happened at the same time. Over a period of a year, you will have your winners and your losers and your winners will, in the longer term, more than compensate for the losers.

Each extra share added beyond 10 has an insignificant effect on reducing the risk.

<!--EZCODE BOLD START--> Ireland vs. Europe<!--EZCODE BOLD END-->

Mith - where do you get the idea that european shares are less risky than Irish shares? Where do you get the idea that very large companies are less risky than large companies? Depending on which period you take, the ISEQ has done well against the Eurostoxx and large companies have outperformed smaller companies. I am not aware of any inherenet reason why the Eurostoxx shoul outperform the ISEQ or why very large companies should outperform large companies.

The only valid objection to investing in just 10 Irish companies is that their performance may be more correlated with each other than say the top European companies would be.

I am not sure that this is as big an issue as people make it out as the top companies are all strongly correlated anyway.

In the particular case of Elan and AIB, their recent problems have nothing to do with the fact that they are Irish.

In Elan's case, if their accounts had been produced under Irish accounting standards instead of US GAAP, their reported earnings would have been lower all along. AIB was the result of what appears to be a fraud in its American arm.

If your entire wealth is in shares, then I would recommend that you diversify overseas - but that would include America and Britain and not just Europe. If your share portfolio is just one part of your overall wealth, then a holding of the top 10 Irish shares in equal parts is fine.

Brendan

valuations

Isn't the problem one of valuation. Many stocks are on stilts. One piece of bad news and they get hammered. Look at Elan.

" sure I know they'll recover and are good long term bets"

Still too much investor complacency!

Is ten enough ?

It would appear not.

Brendan has argued that ten uncorrelated shares is sufficient for diversification. I recall broadly similar data from finance classes long ago.

But intuitively, it seemed to me that while the eleventh share might not reduce 'normal' risk greatly, holding more than ten shares would certainly reduce catastrophic risk (which is what we're focussing on with Elan and Enron, whatever about AIB). Is it as likely that five in fifty companies would go bust as one in ten. Again, intuitively no - not sure about statistically.

In any event, I decided to do some digging. I found a number of articles (links below) which suggest that the volatility of individual shares has increased markedly, and that somewhere between 25 and 50 stocks is needed for sufficient diversification.

www.investavenue.com/article.html?ID=3466
[broken link removed]
[broken link removed]

As I posted on another thread this morning, institutional investors such as pension funds have prudential limits on their exposure to a single stock - generally 3% or 5% - to protect them from such catastrophic loss, and I don't see why individual investors shouldn't do something similar.

As regards Ireland versus Europe, it just seems to me that Europe is much less risky. Intuitively (again) bigger companies are less likely to go bust than medium sized ones. Irish investors have historically had high weightings in Irish equities, but new ones (like the National Pension Fund) are taking a view of Ireland in its global context (ie tiny). Other investors will increasingly do the same. Why limit yourself to companies which have their headquarters in Ireland - that's all it really means. Index tracking via Eurostoxx is, in the aggregate, a lot less risky that investing in the top ten Irish shares, in my view.

Brendan probably feels aggrieved that he's getting so much grief over what may be a once in a lifetime convergence of events. But it has focussed attention on an aspect of his investment recommendations with which a lot of AAM posters were already pretty uncomfortable. I'd be very interested on any more academic/scientific research anyone can recommend.

Re: Is ten enough ?

The question is bigger than simply 'Is ten enough?'. Is it big enough for what??

At one extreme you buy the entire market (via an index/indices of your chosen market/markets) in order to achieve as much diversification as possible. This means your growth will be exactly that of those markets and one or two stocks tanking won't have a significant effect on the entire value. Likewise, one or two stocks significantly increasing in price won't have a significant effect on the entire value either.

If you want your fund to gain more from growth of a smaller number of stocks then you expose yourself more to the volatitlity of those stocks. Nothing new here to anyone investing in the stock market.

So, the question is not 'Is ten enough?' but more 'Is ten enough for you?'. It's not a question of Brendan being right or wrong with his ten stocks, it's just his opinion and his tolerance of risk/volatility and to my mind it's not a bad option in the medium-high risk range.

DD.

I have been trying to get a feel for the level of true risk in holding the top 10 Irish shares. What have been the big disasters over the past 30 years? By a big disaster I mean a sustained loss in value from a previous high. I take 50% as being the loss in value and sustained meaning 5 years. Can anyone add to this list? An Irish Stock Exchange listing from 30 years ago and 20 years ago would be very useful.

Elan is the most obvious recent example, assuming that its drop in value will be sustained.

Eircom is another recent big loser.

Fyffes declined from €4 to around €1 after the .com bubble burst.

Smurfits is probably less than 50% of its high after adjusting for inflation.

Has Waterford Wedgewood collapsed more than 50% from its glory days?
Glanbia/Avonmore is a shadow of its former self.
James Crean is a shadow of its former self, now divided into two shadows of its former self.

PJ Carroll was a quoted company which diversifed from tobacco into mail order and salmon farming and I would guess they lost at least 50% of their value before being taken over.

Power Plc was a property company which went into liquidation - which I think was once in the top 15.

Smaller company disasters include exploration companies, technology companies such as Horizon, Reflex and Memory, The Sunday Tribune, etc.

Most of the shareholders in these companies lost nothing in the long term as they had bought the shares well below the highs. Elan, Fyffes and Smurfits have probably all been good long term investments.

So if you held a portfolio of ten shares for the long term, the spectacular growth from shares like AIB, IAWS, CRH would have more than compensated for any bad luck in the above shares.

The Ten Question

Hi Brendan and DD,

Your posts do raise an important issue - investment context.

I'd agree that ten shares would be sufficient diversification for a high-risk "play money" portfolio for a seriously wealthy investor. But mightn't even one or two shares be "enough" (in DD's phrase) in those circumstances. Diversification isn't really an issue for that type of investment, because the investor's other assets provide more than adequate diversification.

My impression of Brendan's advice was that it was aimed at the typical long-term investor, who can't afford to lose large chunks of his/her money, and whose objective is that it grows as steadily as possible over time. As I said, I think the prudential risk controls taken by large pension funds make sense in that environment. I have argued before on AAM for balanced, well-diversified, portfolios for such investors.

The question I suppose I thought we were trying to answer here, DD, was <!--EZCODE ITALIC START--> what is the minimum number of shares needed to construct such a "well-diversified" portfolio<!--EZCODE ITALIC END-->. I was therefore interested to see that there was some academic research to suggest that that number had increased of late.

But I do accept that investment context and other assets held are important criteria.

diversification

One of my quibbles with Brendan's '10 Irish stocks' advice is that whether or not the number is appropriate, there is a serious overdependence on the domestic economy.
While many of the plcs earn significant revenues outside the State, it is still the mainstay.
An investor whose own income varies in line with the well-being of the Irish economy should probably have a significant part,if not the bulk of their assets exposed to other markets and sectors in which the Irish market is under-represented.
Going back to the number of stocks issue,we now have a number of concentrated funds with 15-35 holdings,so presumably the fund managers believe that is where the answer lies.This is consistent with the academic work.

Interestingly, the last data on the New Ireland Smartfunds web-site (sorry,can't put up the link..)shows that within the Smartstocks 15 fund,25% of the fund was in two stocks,and 36% was in three stocks.Canada Life's Focus 15 seems to have a policy of keeping the stocks at a fairly standard size,while there is no data available on the Target 20 fund from Hibernian's web-site.

Irish Stocks

You can add Crean to the list of catastrophes, Brendan. I can recall it being a top ten stock (maybe ten years ago), and it has since lost over 90% of its value. The property company was Power, I think.

Also don't forget that under your system you wouldn't have invested in anything until it reached the top ten, and even then not till your next periodic rebalance, so you're not getting in at the ground floor by any means on the Elans etc. I suspect (without having the information to hand) that both Fyffes and Smurfit have given a negative return over the past five years, and that Waterford is certainly negative, though maybe not by 50%.

Does any AAM poster have access to the historic information Brendan suggests, I wonder ?

Re: Irish Stocks

Hi Dynamo

Imagine I couldn't remember Power's name! Your right, Crean was a top 10 stock, so I have added it to the list.

That's a very interesting point about not getting in at the lower entry price. Is there a small bias built into my system whereby you only buy shares which are on the up and then watch them fall. I will have to think about that. Elan would have come on to the radar screen a long time ago and went from being worth 10% of the market cap to being worth 25% of the market cap, before its recent fall.

Brendan

Having reviewed Dynamo's links

Hi Dynamo

I studied those three links in some detail and they support me in my views that 10 stocks is enough.

I found Bernstein's article on diversification the most informative.
My big problem with these studies is that they define risk as volatility, so I got excited when I came across the following paragraph in Bernstein: <!--EZCODE QUOTE START--><blockquote>Quote:<hr> This is all profound and important stuff. And, unfortunately, highly misleading. …
There are critically important dimensions of portfolio risk beyond standard deviation... <hr></blockquote><!--EZCODE QUOTE END-->
So I expected to see my idea supported and then he goes on to conclude that you need 200 stocks to minimise risk! He is defining risk as the risk of underperforming the market!!!!
He selected 98 portfolios of 15 stock in 1989 and looked at their returns after 10 years. The S&P returned 19% during those 10 years. So how bad was the worst of the 98 selections? The worst was 14%. Two produced <!--EZCODE ITALIC START--> only<!--EZCODE ITALIC END-->15%.
Coming last in a sample of 98 is a disaster if you are a fund manager. You will get a name for being a dog. But earning 14% a year for 10 years is not a bad result.
A further quote:
<!--EZCODE QUOTE START--><blockquote>Quote:<hr> DIVERSIFICATION: (WSJ) Ronald Surz and Mitchell Price calculated returns for portfolios of 15 randomly selected stocks over the 13½ years through June 1999.
The authors found that among such randomly selected 15-stock baskets, the typical portfolio strayed as much as 8.1 percentage points a year from the market's return. Thus, if the market was up 11% in a given year, the typical portfolio might gain as much as 19.1% -- or as little as 2.9%. <hr></blockquote><!--EZCODE QUOTE END-->

This seems to suggest that in a year, the worst basket underperformed by 8%. I don't think that underperforming the market is a catastrophic consequence.

<!--EZCODE BOLD START--> So to summarise my views:<!--EZCODE BOLD END-->

By <!--EZCODE ITALIC START--> risk<!--EZCODE ITALIC END--> I mean, the chance that something catastrophic might happen. A permanent loss of long-term value would count as catastrophic. Underperforming the market or short-term volatility do not count as risk in my book.
You cannot eliminate risk.
Deposit accounts will probably decline by 50% in real value over the longer term - that is real and substantial risk.
A diversified stockmarket investment will probably beat inflation and the return on deposits over the longer term. <!--EZCODE BOLD START--> I was wrong<!--EZCODE BOLD END--> to convey the impression that there is no risk in such a strategy - but it is the least risky strategy.
Is 10 shares enough? Yes, I still think it is. The market risk is much higher than the stock selection risk from having only 10 shares.
Is an Irish portfolio suitably diversified? I am still thinking about this.

Brendan

Initial Observation

Hi Brendan,

I think you're being too stubborn about this. <!--EZCODE ITALIC START--> Never, they cry.<!--EZCODE ITALIC END-->

You're right - an individual investor has a different motivation to a fund manager, and beating (or matching) the market is not especially important for the former. However, the evidence shows that several analysts who used to believe that 10-15 shares was sufficient for diversification now believe that a higher, perhaps much higher, number is required, and can demonstrate why - higher stock specific volatility. I'll have a think about this and do some more research.

But my one quick initial observation is this. You've selected, via Bernstein's work, one of the best periods for US equities one could find. You state the S&P returned 19% pa over the decade. Let's assume inflation was 4% pa. Stocks thus gave a real return of almost 14.5% pa, or just over twice their long-term expected level.

If you're comparing portfolios against each other, as Bernstein is, that doesn't matter much. But if you're comparing them against an absolute benchmark, as you are, then it does. You'd need to model the data in a more typical period for returns to make your case that even a poor relative performer would have given an acceptable outcome to its owner. Simplistically, if the return from the market had been in line with its expected long-term level, say 11% pa, then the worst portfolios, assuming the same performance gap, would have given a return of around 6% pa. In a poor decade for stocks, and there must be some, to compensate for the excess returns of the 1990s, then a poorly-performing portfolio would do even worse.

Hope this makes some sense. I'll have another look at the academic work, and see if I can convince you !

It is such an important subject

Hi Dynamo

It is a very important subject and I would like to clarify my thinking on it, so I do await your research with interest.

I am stubborn - but only when I am right.

If the index goes down by 50% over a twenty year period, it doesn't matter much which 10 stocks you hold, you are going to hurt. It seems that the range seems to be from -8% to + 15% or so.

I suppose the only real answer is to do some sort of model where you fit in the actual dispersion of returns on large Irish stocks over the past 30 years and see can you produce a disaster from them during normal or bad periods. Disastrous periods are going to be disasters anyway.

Brendan

irish shares!!!

Wheather 10 shares is enough is one question, practicalities favour it though.
But the question of Irish is another one and the more I consider it the less inclined I'm to go that road. Considering some of the benefits that the Irish economy has being enjoying such as the surge in people available and working and the relative benefits of our exchange rate with US and UK, our main trading partners, both of these benefits are on the wane.
Also the present 18% ISEQ weighting of Irish pension funds will be reduced, our they seeking out shares in Small Markets elsewhere in the world? I suspect not so why should we anticipate ISEQ will be sought out.
Given that now we have no currency risk in euro stocks what are the benefits at all in going for the ISEQ apart from the advantage of being able to hold your own share certificates?

Irish ?

An article in the fund management supplement to the FT today deals with the "home or away" question .

Reporting the findings of a study by Michael Schroder of the Centre for Economic Research(Mannheim), it suggests that depending on the domestic market's profile, a significant improvement in the risk-adjusted return can be achieved by diversifying internationally.

His analysis suggested that a UK investor would improve his RA return by having just 10% at home and that Germans should forsake their home market completely ! At the other end he finds that US investors should stay at home.

If this is valid for large markets like those, the notion of the 10 stock Irish portfolio seems even more peculiar.

this idea

There seems to be a vacum in that area where once there were recommendations for the Buy 10 Irish shares. Is it safe to say there's no support for this idea anymore.

Re: this idea

Hi Aimless

I think there is a difference of opinion. Let me summarise my opinion

There is some risk that equities are very overvalued at the moment. This market risk is much higher than the extra risk in buying 10 shares instead of buying 20 shares. In other words, people reading this thread might think that buying 20 shares is much less risky than buying 10 shares. I don't believe that it reduces the inherent risk in equities significantly. If you are worried about the risk of 10 shares, then you should avoid equities altogether.

I recommended buying Irish shares because they are easy to buy and keep an eye on. Buying European shares directly has been a nightmare for me. Buying American shares is probably easier.

If you are not buying shares directly, it is as easy to buy a Eurostoxx tracker as it is to buy an ISEQ tracker.

The ISEQ has a big part of its total value in the 4 top shares, so an ISEQ tracker is probably too concentrated.

I wasn't making any judgement that the Irish stockmarket is better or worse value than the American or European stockmarkets. Having recently read Shiller's Irrational Exuberance, I am now firmly of the opinion that the American market is significantly overvalued. The Irish market seems to be reasonable value and the Eurostoxx seems fair as well, but mainly because it has dropped by 50% from a very high level.

There are a lot of telecoms stocks with very high price earnings ratios in the Eurostoxx, which makes me nervous. But if you want to diversify outside Ireland and if you are buying a tracker, then the Eurostoxx is a good alternative to the ISEQ.


Brendan

I think the basic advice to buy teh top ten Irish shares ten shares and leave them is risky (i.e. dangerous) advice especially for the first time investor. ("We recommend that you buy a portfolio of the top 10 Irish shares and hold them for the long term.")

Apart from havng to decide exactly what 'top' means (e..g top in terms of earnings, capitalisation, dividends, whatever, etc.?) not all shares will increase in value in line with the stock market. To take an example from a book for armchair investors, a pick of blue chip shares declined by 7% in a period when the FTSE 100 rose by 8.2%.
Surely unless you do some fundamental analysis and have developed an investment strategy just buying 10 shares (even the 'top' ten) and leaving them is risky (i.e. ?risk? in the sense of loosing all/most of your investment). Even if you buy in equal money proportions the top ten shares in the ISEQ, you have to re-balance your portfolio each year. You just can?t leave them.
Also most ?buy ten shares? strategies are just a variant on establishing your own (possibly non-weighted) index. If you do do not want to be a stock-picker why not just put your cash into QL?s Celtic Freeway, which is an index based on the top 20 ISEQ shares? You have low running costs and the share selection is done by the market. Why go to the cost of buying ten shares unless you think that you can beat the market?
If putting all you money in a QL tracker (or any tracker) is too risky (i.e. volatile) you could split it say 80/20 with a bond fund? The point I am making is that unless you are a stock-picker and have an investment strategy you might as well just invest in a tracker (or trackers), balanced with bonds if you are risk adverse, or to balance volatility.