How You Can Beat Fund Managers By Cloning Their Best Ideas

16th Jan '14 by Edward Croft

19 comments 17677 reads

It's increasingly common knowledge that actively managed funds significantly underperform the stock market. The level of underperformance varies in different market cycles but on average it is of the order of 3% per year. Given this, is there really any point in paying attention to what fund managers are buying?

Well the latest cutting edge research shows that there is. Fund managers are actually way better stock pickers than these numbers suggest. By separating the wheat from the chaff in their bloated portfolios you can find their very highest conviction ideas... ideas which dramatically beat the market, by up to 20% per year !

So what are a fund manager's 'best ideas'?

In hunting for their best ideas the first place to look would be amongst their top 10 holdings, but this would be naive. A fund's performance is normally assessed relative to a benchmark index like the FTSE 100. If a fund underperforms this index the investors often pull out their money and the fund manager can lose his job.

So to manage this career risk, the cautious fund manager often weights most of the stocks in his portfolio similarly to the benchmark, with proportionally larger weights in the stocks weighted more heavily in the benchmark. It's only for his very best ideas that he'll deviate these weightings most substantially.

Indeed a ground breaking 2005 research paper called "Best Ideas", Randy Cohen, Christopher Polk and Bernhard Silli researched this common sense idea. For every fund in the market, they calculated the deviation or 'tilt' of each holding from the benchmark weight. Portfolios constructed using the top 25% highest conviction positions held by fund managers showed anywhere up to 23% excess annual returns depending on the technique used!

How can you maximise the returns from high conviction fund manager ideas?

We won't go into the precise technique used in the paper as it's beyond the scope of this article, but there are several key findings from the paper which can significantly improve the results from cloning best ideas:

Stick to each manager's top pick - The stock with the very highest 'tilt' for each manager outperforms the rest of their picks. Descending down the rankings through the second, third & fourth best ideas 'monotonically' reduces returns. When assessing the strength of a fund manager's idea it's worth bearing this in mind. Anything outside the top five best ideas is unlikely…

Unlock with Email

Unlock with Google

Already have an account?
Login here

Fangorn

16th Jan '14

Interesting opening comment Ed

"It's increasingly common knowledge that actively managed funds significantly under perform the stock market."

So the solution would be to buy a tracker surely?

But I recall reading that Active managers consistently outperform the trackers apparently so where does that leave us?

Active funds trounce passives over all time periods
http://www.trustnet.com/News/482056/active-funds-trounce-passives-over-all-time-periods/2/1/

The rest of the article was a superb read however.. Thanks for scribbling.

jjis

Until it launched (don't suppose there is a firm date for that yet Ed?) people might be able to find a few ideas from here.

bsharman

good article Ed, which I have just read on my train home. There are 'alpha' managers who outperform over long periods such as Antony Bolton (pre China), Giles Hargreve, Harry Nimmo, Mark Slater etc. I like to follow them because they have a proven record over 5, 10, 20 years. the difficulty is finding out what they are buying - outside of their top 10 holdings, which will change monthly and which everyone will have access ie most of their top 10 have already grown many times over hence why they have reached the top 10..

Edward Croft

Firm date - pah ! It was supposed to be December. The data is taking longer than we'd have liked as we have grown to understand the database & partly because we keep getting better product ideas.

From your link - these massive FTSE 100 positions are unlikely to come out as the best ideas in our analysis. The Standard life guy looks to have much more interesting positions than the others - probably a smaller fund.

I think Giles Hargreave's fund has grown way too big and his positions too small for his future performance to be as good. Slater still does a lot of conviction picks. Big weights in small caps. I respect that a lot and the research shows that high conviction managers generally outperform. Of course the Slater Zulu strategy is doing rather well here - or rather better than Slater himself :-)

We are mostly focused on what purchases are happening recently and in size. That's where most of their insights are.

So the solution would be to buy a tracker surely?

Trackers have been pretty effective so far, but I've written extensively about what I see as their issues. See "Index Funds are Parasites and are going to kill the market". I'm not the only one who believes that the growing dominance of trackers & ETFs is potentially going to become a mis-selling issue in the market, or lead to a gross mis-allocation of capital.

Beyond this, by definition if you are buying a tracker you are buying a huge number of worst ideas. Why not focus on the very best stocks with the best profiles? Our work here is designed to focus on what works in markets... quality, value and momentum ... these are some of the core driving factors in the stock market (as verified by reams of academic studies), but beyond this of course there are always great stock picks that don't have this profile. The ownership / smart money add-on we are building is designed to try to capture these additional sources of ideas.

I see nothing wrong with having trackers as a good part of a portfolio as they are a low cost solution for getting equity ownership. I just know that people can do better with even a tiny amount of elbow grease.

jjjs - this is interesting and I hadn't seen this before, but I do think they've got their ranking wrong.
http://citywire.co.uk/money/everything-you-wanted-to-know-about-top-stocks/a522114

Citywire are generating conviction ranks from consensus, which is completely opposite to what the research shows. The best ideas are when they are best, fresh and first. When top conviction ideas are shared across multiple managers they tend to underperform.

Great that they've done this though !

[edit]

On second read I realise they are only ranking 5 fund managers ! That's such a tiny universe and there would be v. v. little alpha in it imho. We are looking at tens of thousands of funds to find the knockout positions - including hedge funds and international funds... am sure there will be noise, but it will be pretty fascinating - the output is already extremely eye opening. I've come across loads of conviction small cap funds I'd never heard of like this one which has some seriously big bets in it http://hcefurb.com/

Agree about the FTSE names, if you scroll down the page it gives a better idea of the conviction picks in descending order.

A better alternative to Cap Weighted Trackers might be the Smart Beta or Fundamental Index Funds with weights based on Valuation Factors. A good discussion of these and the pluses and minuses appeared in the Economist last year.

21st Jan '14

Ed,
"Beyond this, by definition if you are buying a tracker you are buying a huge number of worst ideas. Why not focus on the very best stocks with the best profiles?"

"I see nothing wrong with having trackers as a good part of a portfolio as they are a low cost solution for getting equity ownership. I just know that people can do better with even a tiny amount of elbow grease. "

Couldn't agree more with those two statements...

"The ownership / smart money add-on we are building is designed to try to capture these additional sources of ideas."

Sounds interesting, as indeed were comments I read somewhere that Live Prices are coming to Stockopedia possibly as well.

We do get asked for new features all the time but can't build them all at once ! Live prices we'll definitely do in future, but not in 2014. This year is one for the ownership database, USA, and some other goodies up our sleeve. Will keep you informed as the year goes on.

22nd Jan '14

Fully understand. It's an amazing job you guys have done from a standing start. USA sounds interesting - to be similar to the UK set up but with a separate sub?

Grateful for the updates. Watching with interest I have to say.

marben100

Hi Ed,

Not sure if you're aware that the HCE FURB is run by Guy Thomas, author of "Free Capital". Not surprising you find some good ideas there. ;0)

Cheers,

Mark

Ahhhh - you know I did suspect as much !

timarr

24th Jan '14

I agree absolutely that there is an indexing bubble. There's evidence to show that large caps have been systematically overvalued due to the passive effects of market cap indexing. It's possible that last year's underperformance of index trackers compared to active funds was due to this - the article indicates as much. But given that many active funds are closet trackers these days it's just as likely to be down to the fact that everything has gone up, so simply tracking the index and having a few conviction bets on the side will have ensured outperformance. A rising tide and all that.

And, even if not so, I'd be quite careful not to draw the conclusion that active funds are better.Long-term academic studies show that active funds don't add alpha, that a randomly varying two-thirds of them underperform index trackers each year and that even when people do pick the right active funds they tend to trade away their gains by buying and selling at the wrong time: basically active fund performance tends to mean revert and people tend to chase returns.

There are a few interesting snippets in there though, because as the article shows, the effects of charges on passive funds can vary massively. Given that two FTS100 trackers are much of a muchness you'd expect the one with the higher fees to go out of business or adjust its prices. But nope, they carry on raking in money from gullible investors based on brand and marketing and sheer lack of financial education.

But the lesson I'd draw here is that index trackers are right for most investors, but generally people would be better off in a mix of tracking funds, varied across geography and asset class, rather than focusing on a single market cap weighted index.

Absolutely agree timarr. In fact your piece on the index bubble is one of the most terrific on the web. "Modern finance is a witches brew indeed" http://www.psyfitec.com/2012/03/hubble-bubble-index-trouble.html

Have you got any other links from your blog that you could post here on this topic?

PS - everyone should read timarr's Psyfitec blog. http://www.psyfitec.com/ We do syndicate pieces regularly but all the best stuff is there.

cig

Well given recent small cap outperformance, isn't all or most the alleged overvaluation gone now?

And in any case the potential impact is limited. Worst case scenario is that 100% of listed equity capital goes to large cap indices' members, but they still, due to the natural power law distribution of company size, will represent 80% of underlying real world listed assets (measure it by fundamental metrics if you think market prices are wrong). So the max possible overvaluation is like 20%, just because small caps are, well, small, and if we got there ex-index companies could be had for a song (I wish).

I think that an index bubble is more likely to manifest itself in increasing idiosyncratic volatility (as fewer people would price individual stocks) than in overvaluation. Would be great for stock pickers if it was true...

Mark Simpson

I'm not convinced by the argument that popular indexes are always a 'bubble'. I believe indexing can have a big impact but since it's the flow of money that determines prices rather than the absolute amount it's more likely to increase volatility rather than absolute value. Particularly since the average market participant is terrible at market timing - taking their money out after big index falls and putting it in after big rises.

I also think that the index arb is so well known that these days it isn't index arbs taking money off naive index investors but hedge funds taking money off naive index arbs!

Of course one of the benefits of owning and market-cap weighted index compared to a fundamentally weighted one is that it re-balances itself without trading costs (which seems to be the reason why the fundamentally weighted indexes have often failed to deliver the promised theoretical out-performance.) The effect of this is that relative value anomalies within the index remain unaffected by index flows.

That the small cap indices have significantly outperformed the FTSE100 in the UK last year suggests to me that the FTSE100 is unlikely to be a 'bubble'. Increasingly I'm finding value in the unloved sectors of the FTSE100 than in small cap shares. Buying FTSE100 shares is different to buying small cap shares though. When buying a small cap you are looking for where a large seller has indiscriminately sold down a share to an oversold level and liquidity or under-research means that this pricing anomaly has not been corrected yet. With large caps pricing errors are unlikely to be caused by liquidity or under-research. Instead mis-pricings tend to occur because sectors become unpopular (e.g. supermarkets?) or due to fund manager incentivised behavior (e.g. selling losers at year end so they don't appear in the annual report, or preferring high-beta stocks as a proxy for leverage in long-only funds.) The passive role of index funds means that these anomalies remain for those who are unconstrained or willing to take a longer term view.

Ed,
Quick question given the debate on here re active funds v trackers. Presume your data is going to be looking at OEICS. Some suggest that Investment Trust have generally out performed although this may be down to lower charges and gearing. However, the closed end structure does allow their managers to be more long term / have more conviction for example the two I wrote up recently on my Compound Income? blog. Any thoughts of including IT's in your data or maybe you are?

How you can beat fund managers by cloning their best ideas

So what are a fund manager's 'best ideas'?

How can you maximise the returns from high conviction fund manager ideas?

Unlock the rest of this article with a 14 day trial

About the Author

Edward Croft

19 comments

Unlock the metrics for all these stocks and more...

Unlock the metrics for all these stocks and more...