What’s so special about investment trusts – and some of their dangers

Peron pointing at a blackboard © Getty Images
Investment trust outperformance: long suspected by investors but now proven by boffins (NB no science was performed in the making of this picture)

What if I told you there was a type of investment vehicle you could have bought 18 years ago that would have beaten the type you mostly own by about 1.4% per year every year since? You’d do the sums pretty quickly (18 years of that scale of compounding adds up), and my guess is that you’d then want to own it pretty quickly too. Good news: you can have that vehicle.

According to research by Professor Andrew Clare and Dr Simon Hayley of Cass Business School, it is investment trusts (also known as closed-ended funds). These are listed companies, the business of which is investing.

Upon listing, they issue a set number of shares that trade and can be bought and sold just like any other share. Unlike the more popular kind of fund in the UK – open-ended funds – often called unit trusts or open-ended investment companies (Oeics) – which constantly issue and cancel units based on investor demand – investment trusts have a set pool of long-term capital to invest.

Why investment trusts outperform unit trust and Oeics

It has long been suspected that closed-ended funds outperform their open-ended counterparts, but these numbers are particularly fascinating. That’s partly because it is the first study of this type to be done by neutral academics rather than industry supporters, but mainly because the gap is way bigger than past studies have suggested (usually about 1%).

Cass did a little work to try to explain the difference. Some of it is down to sector bias – investment trusts have had more of a slant to smaller companies than open-ended funds (and smaller companies tend to outperform).

There’s also a small impact from gearing (investment trusts can borrow money to invest, and this can have a positive influence on returns) and from share buybacks (which reduce the number of shares in issue and so push up the value of the remaining shares).

Then there is a small effect from “survivorship bias” (bad funds die fast). But even after you have added up all these effects, say Prof Clare and Dr Hayley, 0.84% per year of performance remains unaccounted for. As all compounding obsessives know, that’s a difference that adds up to real money.

The next question then has to be, what did Cass not take account of that could explain the rest? One factor is governance. Investment trusts have boards whose job is to represent the interests of shareholders. For disclosure, I sit as a non-executive on two investment trust boards, Baillie Gifford Shin Nippon and Montanaro European Smaller Companies.

You could use share buybacks as a proxy for a board’s success (boards decide when to do these and in what scale). But there’s more to it than that. Good boards keep admin costs consistently low. They keep the managers on their toes (it’s all too easy for fund managers to forget that they are the service providers to investment companies whose assets they look after, not the owners of those assets). And they can bring valuable experience to the investment process.

There are fees. Cass used net-of-fee numbers for its research, and the fees charged on investment trusts have historically been lower than those of open-ended funds. I reckon that could account for 0.3% (ish) of the difference.

But the really interesting – and utterly unmeasurable thing – might be manager mindset. Does knowing that they are working with permanent capital help the managers of investment trusts make better long-term decisions?

Being academics, the report’s authors can’t be too enthusiastic about this, but they do headline their research with the phrase “Investment Trust structure may be more conducive for active management”.

Some dangers to be aware of with investment trusts

So what’s the catch? When you have lots of data, you have lots of caveats – but the key one is this: these numbers tell us much about the past. They tell us less about the future.

As the authors make clear, you can can’t say that investment trusts outperform. You can only say that they have outperformed, which is different – and awkward given that all you want to know is what will happen in the future. I can’t read it any more accurately than anyone else (as regular readers will know).

But there are a few factors to consider. First, investment trust fees aren’t lower any more. They are at best on a par with OEF fees, and often more expensive. Second, issuance. Last year saw the highest levels of issuance in the investment-trust sector for ten years (up 77% on 2016). Not all of those funds will survive. Third, gearing. Cash is cheap and boards have started to borrow long-term (the Scottish Mortgage Trust has just taken out a 30-year loan). If markets keep going up, all that borrowed money will quickly pay its way. If they do not, it might do the opposite. Let’s not forget that investment-trust outperformance has come against a background of generally rising markets.

Finally, discounts. These are not discussed in the Cass paper (understandably so, as the authors were mostly attempting to measure managerial skill and compare like with like). That meant looking only at the net asset value of investment trusts, not at their share prices – which can diverge hugely from the underlying asset value represented by each share.

But in the real world, for real investors, the shift in the discount (or premium) to the net asset value can make a major difference. Let’s say you buy a share at a 5% premium to its 100p net asset value (so 105p). The market falls and the net asset value does too, to 80p. At the same time the premium disappears and the shares start trading on a 10% discount, so 72p. You haven’t lost just the 20p in net asset value, you’ve lost 33p. Ouch.

That risk is worth paying attention to. According to Winterflood, the average discount across the investment trust sector has been 9.4% since 1989. Today, it is 4.2% – something that suggests an about-turn ahead. You can make a case for discounts being permanently lower if you try – more boards have put in place discount control mechanisms and aggressive buyback policies for example – but, generally speaking, ignoring such firm market messages can be a mistake.

So should you be looking to hold more investment trusts in your portfolio? Even with all the worries about discounts, fees and leverage, my answer would still be yes. We can’t explain away the full outperformance of the past without constantly returning to the idea that something in the investment trust structure makes managers better. That doesn’t make investment trusts magic. But it might make them special.

• This article was first published in the Financial Times