"Inversely correlated funds"

valiant24

I'm in my early-to-mid 60s, not really working any more, living off my and my wife's SIPPs, ISAs and dealing accounts.

I've been a disciple for some years of Lars Kroijer and others, holding largely just global trackers and cash/gilts in a 70/30 proportion, plus one fund (Fundsmith) which is now so valuable I can't sell it off because of the CGT!

I do like the writings of Sir John Kay though.   Kay, who's 76, writes that he keeps virtually no cash and doesn't own bonds, and has his money invested mainly in global trackers, but diversified against other equity funds which he judges to be likely to be inversely correlated with global trackers.

I've never read him name any though, and am pondering what he might mean.  Maybe so-called "wealth-preservation" funds like Capital Gearing Trust or Ruffer?

Any insight or just opinion gratefully received.

InvesterJones

For an equity (as opposed to alternative asset) fund to be inversely correlated with a global tracker it would basically need to be concentrated in stocks that don't correlate with the stocks making up the majority of global trackers weighting. So if the global tracker had a high proportion of technology stocks then you can bet against the market by picking up some other kinds of stocks, like utility or miners.

JohnWinder

I can't find him writing on that, but he's an engaging writer with plenty to offer. Some of it is puff to fill a column, eg he pushes readers to hold individual shares, 'might want to consider' he says. I think it's well established holding an individual stock is not the best approach, although he's not saying 'do it' or even 'think about doing', rather he's saying 'you might think about doing it'. That safely removes him from criticism, as well that the suggestion is after you've mostly invested in a diversified fund, so not much harm to be done with a single stock; but if you only risk a small holding on a single stock it can't do you much good even if it's a winner. It's entertaining writing but will that sort of idea help us?

I can't find your 'inversely correlated' reference, perhaps you could share the link, but other mentions of correlation are a superficial consideration. Firstly, correlations between assets varies over time, so good luck predicting what it will be in future which is when you need it to be whatever you want it to be. Secondly, uncorrelated is useless if the return is poor. One could hold a SP500 fund and another which shorts the SP500, and you'd have nicely uncorrelated assets but it wouldn't give a good return, I think.  Further discussion here: https://www.bogleheads.org/forum/viewtopic.php?p=6716038&sid=ceb87516a0dce450cf64a1c7d0940d0d#p6716038 

And: https://www.bogleheads.org/forum/viewtopic.php?t=181077

https://www.bogleheads.org/forum/viewtopic.php?t=266366 here's an attempt to quantify the issue.

https://www.bogleheads.org/forum/viewtopic.php?t=371800

LHW99

I suppose if the Global Tracker is Developed World, then there is an argument that an Emerging Markets Tracker would be (at least partly) uncorrelated, although many countries in that category tend to mirror the dollar.

Or ig the tracker is Global Large Cap, then a Small Cap equivalent could be an addition, although that's mainly diversification.

I do look at a basic x versus y graph for some pairs of the funds / IT's I use, and check the Excel R^2 value between the unit prices, although that I'm sure isn't eaxactly what is meant!

Malthusian

Why would I want a fund that goes down over the long term? 

Absolute return funds, life settlement funds, Greek shipping, and withdrawing some of your money from the bank and setting it on fire are all asset classes "inversely correlated with global trackers" which have been very popular in the past.

"Capital preservation funds" like Ruffer should in theory not fit the bill, if they achieve their aims; those are supposed to go up over the long term but more slowly.

talexuser

If you realy want to inverse track, just look up the geographical makeup of the global, eg % of USA, China, UK, Germany etc etc and buy a proportion of country or sector trackers in the inverse %s?
It would be interesting to see what the result would be compared to the global in the event of a fall led by USA or China, once the dust has settled and other areas have recovered somewhat by realisation they won't be as badly affected.

Malthusian

That wouldn't be an "inverse tracker". For an inverse tracker you would need to short those markets in proportion to how they make up the global index. Which would rapidly lose you money until you got margin called or cut your losses.

An opposite global tracker might look something like USA 32%, Europe 6%, Japan 9%, Pacific 13%, UK 13%, Emerging Markets 27%. I had to fudge the numbers a bit because you can't hold 97% in emerging markets, 96% in the UK, etc etc.

That would have underperformed the global market in recent years, because the global market is heavily weighted towards the US, which happens to have outperformed the global market since the dot com crash. But not by much; the opposite-day portfolio would be up 9.2%pa over the last ten years while the FTSE All World is up 11.4%.

If you have a basket of index trackers which have mostly gone up, then a basket of the same trackers in different proportions will still go up, just not by the same amount. For inverse correlation you would need to short those markets. 

If you have a basket of index trackers, some of which have gone up and some of which have gone down, then holding them in different proportions could result in inverted performance. But that's not how global equity markets work. They are positively correlated with each other in the long term. It's not a zero sum game where the US can only go up if China goes down.

EthicsGradient

Are we sure that John Kay advises looking for inverse correlation, or just lack of correlation? For instance, from johnkay.com:

The conventional investor has improved returns by minimising charges. The intelligent investor plans additionally to benefit from market inefficiencies by making judgments. He or she will diversify more than the conventional portfolio allows; that portfolio is heavily biased towards the stocks of large global companies all of which sell to similar markets. A market index contains many stocks but that is not the same as diversification, which emphasises lack of correlation between the likely outcome of different investment choices. And the intelligent investor has the opportunity to stand back from the waves of optimism and pessimism which drive markets — and to avoid bonds, whose current yields should be of no interest to any long term investor.

How to be your own investment manager - John Kay

(For clarification, that was written in 2016, so "current yields" of bonds does not mean today. I've no idea what he currently thinks about them.)

Looking at an X-ray of my holdings, it tells me that (unsurprisingly - others have predicted it above) the correlation (over 3 years) of an emerging markets index tracker with a global tracker is quite low, with a Europe tracker closer to the global, and a US tracker closer still. But this is not just about overall return - I also have a biotech IT that is reasonably well correlated with the global tracker, but it has lost money, where the trackers have more or less stayed level or gained.

talexuser

Malthusian said:

An opposite global tracker might look something like USA 32%, Europe 6%, Japan 9%, Pacific 13%, UK 13%, Emerging Markets 27%.

That's more or less what I was tougue in cheek suggesting, maybe with less USA, but interesting your conclusions on how it would have performed.

valiant24

EthicsGradient said:

Are we sure that John Kay advises looking for inverse correlation, or just lack of correlation? For instance, from johnkay.com:

The conventional investor has improved returns by minimising charges. The intelligent investor plans additionally to benefit from market inefficiencies by making judgments. He or she will diversify more than the conventional portfolio allows; that portfolio is heavily biased towards the stocks of large global companies all of which sell to similar markets. A market index contains many stocks but that is not the same as diversification, which emphasises lack of correlation between the likely outcome of different investment choices. And the intelligent investor has the opportunity to stand back from the waves of optimism and pessimism which drive markets — and to avoid bonds, whose current yields should be of no interest to any long term investor.

How to be your own investment manager - John Kay

(For clarification, that was written in 2016, so "current yields" of bonds does not mean today. I've no idea what he currently thinks about them.)

Looking at an X-ray of my holdings, it tells me that (unsurprisingly - others have predicted it above) the correlation (over 3 years) of an emerging markets index tracker with a global tracker is quite low, with a Europe tracker closer to the global, and a US tracker closer still. But this is not just about overall return - I also have a biotech IT that is reasonably well correlated with the global tracker, but it has lost money, where the trackers have more or less stayed level or gained.

Yes, it was this book.   I don't have it to hand now, but you're probably right: most likely it was lack of correlation that he urged.