The woefullness of market-cap based trackers

Porcupine

Over on the H-L charging structure thread there was a bit of discussion about trackers based on things other than market cap.

There was a FT/Moneyweek article which refers to some interesting research. Despite what MoneyWeek says the research isn't really that hard to read if you skip over the citations and the bits where they describe their models in detail.

I thought I'd share this figure:


Over the past 40 years, taking ten million 'monkeys' and asking them to the pick an index at random, 60% of them beat an index based on market cap.

So it seems indices based on market cap are a really bad idea.

Additionally interesting is section 6 of part 2 where they consider a mechanical 'sell if it's going badly' strategy - if in the last month the market went down compared with the 10 month average, sell and buy treasury bills. If it went up, buy the market. That seems to have the same returns as staying invested over the long term, but less volatility.

Food for thought...

Glen_Clark

Porcupine wrote: »

if in the last month the market went down compared with the 10 month average, sell and buy treasury bills. If it went up, buy the market..

Sounds like a recipe for buying high and selling low?
May have contributed to the dot com lunacy of 2000 - index funds had to buy these dot coms however overpriced they were.

racing_blue

That's interesting but as always with statistics, the Devil is in the detail & can be presented in a way that suits the author.

For exampler, if 40% of the time monkeys beat the market by 1%, but 60% of the time lose by 0.5%, it would still pay to be a monkey. But of course, if I was a fund manager, I'd just say "monkeys lose 60% of the time".

Not saying this applies here & have not had chance to read your link yet - some bedtime reading this evening I think!

gadgetmind

Monevator's take here.

http://monevator.com/why-market-cap-investing-still-works/

I use mostly passive investing but have added some "sauce" to this by over-weighting small/mid cap, EM/Asia, value, etc. Some of these satellites use active investment.

redbuzzard

Porcupine wrote: »

Over on the H-L charging structure thread there was a bit of discussion about trackers based on things other than market cap....

Over the past 40 years, taking ten million 'monkeys' and asking them to the pick an index at random, 60% of them beat an index based on market cap.

So it seems indices based on market cap are a really bad idea.

Additionally interesting is section 6 of part 2 where they consider a mechanical 'sell if it's going badly' strategy - if in the last month the market went down compared with the 10 month average, sell and buy treasury bills. If it went up, buy the market. That seems to have the same returns as staying invested over the long term, but less volatility.

Food for thought...

Hmm.

I wonder how realistic the simulation was. If we were all monkeys, on average there'd be the same amount of money chasing a £10m value company as there was BP.

It's the fund managers who need to deal with the monkey with a pin proposition, not the trackers.

In a way, it's not surprising.Very large companies have stronger dividends, but lower growth potential (generally). The 250 tends to do better in a rising market than the 100. Is it related, in terms of the distribution of growth vs. income across the market?

What about volatility? I don't hold IP Income because I expect to beat a FTSE AS tracker, but because I hope for something close to that with lower volatility and a better compensating yield when the market is down.

The HSBC FTSE tracker I bought on 18/6/2009 was at 85.2% total return yesterday, the best performer I bought that summer when I started a SIPP with a lump sum.

It doesn't say much for other, actively managed, choices if the monkeys are better than that.

Perhaps the worst that can be said of trackers is that they can distort/create the market, as with the dotcom phenomenon mentioned. I reckon the monkeys would definitely distort it more.

JohnRo

What about volatility? I don't hold IP Income because I expect to beat a FTSE AS tracker, but because I hope for something close to that with lower volatility and a better compensating yield when the market is down.

I've bought IP distribution for the same reason, I know it won't make me rich, the intention is for it to provide a stable foundation. Whether that proves to be the case though remains to be seen.

redbuzzard

JohnRo wrote: »

I've bought IP distribution for the same reason, I know it won't make me rich, the intention is for it to provide a stable foundation.

Me too. "Banked" some of the pure equity outperformance into it, and into GARS, which might be finding its mojo again with any luck.

grey_gym_sock

IP distribution will make you rich if you invest enough cash in it, for long enough

equal-weighted is a very plausible approach if you're buying individual shares yourself; but much less plausible for an index fund.

the conclusion should be that there's some case for having a small-cap bias, and a value bias.

also see: http://www.rickferri.com/blog/investments/no-free-lunch-from-equal-weight-sp-500/