Size matters: how small is too small?

In my free time, I like to kick back and relax by reading books with titles like ‘All about Asset Allocation’ and ‘The Intelligent Asset Allocator’. Unsurprisingly, based on their titles, most people would prefer to watch their company’s compliance training videos – the ultimate test of anyone’s boredom threshold – than read these things. These books are for the investing nerds.

I was reading one such book recently (not one of the ones mentioned), when the author recommended something I thought seemed bizarre. He was suggesting investors allocate a 1-3% position to gold.

Now, I’ve made my thoughts on gold pretty clear in this post: ‘Should you own gold as a UK investor?’ – I don’t like it. But that wasn’t what I thought strange about this recommendation. I’d be equally perplexed if it was any other asset class.

I was simply thinking “What’s the point in allocating 3% to anything?”

A 3% weight to an asset class isn’t going to make any real difference to your portfolio. It’s just not large enough to move the needle.

And that got me thinking: what’s the smallest allocation I’d be willing to make in my portfolio? 1%? 5%? 10%? More?

As usual, I consulted my semi-organised collection of accumulated Word document notes, but they came up dry. I hadn’t read anything over the last 5 or so years which had tackled this question.

So I thought I’d see if I could put some numbers to it myself.

Minimum position sizes

Off to Excel we go.

The first port of call was to have a look at the impact on an overall portfolio if a single holding rises in value given varying position sizes:

As an example, the table shows if you hold a 1% position in something (gold, bonds, stocks, property, whatever), then if that holding gains 50%, your portfolio as a whole will gain 0.5% if the rest of the portfolio remains flat. If you hold a 5% position in something, and that something gains 100%, then that will add 5% to your portfolio’s returns.

Different people will draw different conclusion from the table. Beauty is in the eye of the beholder in terms of what you consider a ‘significant’ impact on your portfolio to be. For me, the numbers only really start getting interesting when  position sizes move above 10%. For positions smaller than 10%, you’d need to generate some pretty heroic returns (100%+) for that position to have any real impact on performance.

When I first put the table together, I was thinking as a shorter-term investor. “If I put 5% into crypto, then how soon can I buy my Lamborghini?”

The answer: not soon at all. Even if it doubles, that’s only a 5% boost to my portfolio.

But we should be thinking like long-term investors here, and thinking about how we should allocate our long-term holdings – not short-term speculative positions. We should therefore be viewing the return figures on an annualised basis.

Most assets we’ll be investing in will have annualised returns somewhere between 0-10% – likely closer to 5%.

If you cast your eye down the 5% column then, the figures can now be thought to represent the annualised increase in return you’d achieve with various position sizes. For example, a 1% strategic allocation to an asset class which rises 5% per year is going to add 0.05% to your portfolio’s annualised returns. In other words, a 1% allocation is pointless.

A 5% allocation will add 0.25% a year – not much, but getting better. There are still likely other areas of your financial life which are going to move the needle more.

A 10% allocation will add 0.5% per year. Now things start to get interesting. At 0.5% per year and above, these numbers start to have a meaningful impact on portfolio sizes when compounded over an investing lifetime.

At a 10% position size, long-term returns of 3%-8% (which is the range of what we can expect from equities) will add between 0.3%-0.8% per year in returns. Not mind-blowing, but we’re considering absolute minimum allocations here.

Still, the more you allocate to an asset class (i.e. the more confident you are in its ability to generate its expected returns), the larger the effect on overall portfolio returns.

As I mentioned earlier, this is pretty subjective stuff. My personal takeaway from the table is that a long-term allocation of anything under 10% isn’t worth getting out of bed for. But if you believe an asset class has a much higher long-term return potential (e.g. above the 5% a year I’ve been using as a yardstick), then smaller allocations can have equally large impacts on your portfolio.

Diversifier position sizes

The previous table is useful for thinking about any asset class, but really focuses on what happens when you take a strategic position in a long-term growth asset, and it goes up. Like equities.

But it’s also interesting to look at position sizing in your diversifiers – your safe assets used to buffer the portfolio when the excrement hits the fan, and your risky assets start falling. In fact, this is probably where I see small sizes most often recommended – like our author recommending a 3% position in gold.

So I took a look at what happens if equities fall 50% – which they’re liable to do every decade or so – at varying levels of diversifier allocation levels and returns. (The table assumes there are only two holdings in the portfolio – equities and the diversifier)

As an example, if equities fall 50% and an investor holds 99% in equities and 1% in a diversifier, then the portfolio will still fall between 49% and 50% – even if the diversifier gains as much as 50%.  

At a 95%/5% split equities/diversifier, you’re still not getting much relief from the equity drawdowns, even if your diversifier rises by 50%.

We saw from this post: ‘What’s the best crash protection for your portfolio?’ that traditional diversifiers haven’t tended to rise more than 30% during the post-1985 selloffs. And going back further, this post: ‘Have bonds ever failed?’ shows US 10-year bonds have never gained more than 30% during a crash – even during the great depression.

And that makes sense – most diversifiers are chosen because they’re safe. They don’t go up much, they don’t go down much – they truck along providing small positive returns most of the time, then provide higher returns when risky assets fall. But they’re less volatile by design, so we can’t expect them to provide gigantic returns during a crash.

Looking at the table, if your diversifier rises 30% when equities fall – which is still a big ask – then it’s not going to affect your portfolio’s drawdown by any meaningful amount unless you hold at least 10% – reducing the drawdown from 50% to 42%. Ideally, you’d want to own at least 20% – reducing the drawdown to 34%.

Now, I’m obviously thinking about bonds here, but there are plenty of other diversifiers on the table. There’s gold, puts, long-volatility strategies, as well as every ‘alternative’ fund under the sun (market-neutral, long-short, total return, defined return, buffer funds, etc). Many of these advertise themselves as being able to provide outsized returns during a crash. Tail-risk hedging strategies, for example, are designed to lose a small amount every year, but then shoot the lights out when markets fall. These strategies, if they work as intended (and that’s a mammoth ‘if’) are able to get away with occupying a smaller space in your portfolio, as their returns during crashes are supposed to be large enough to make a difference even with a relatively small allocation.

But even if you picked an asset which was god’s gift to diversification and rose 50% when equity markets fell (such as a tail-risk strategy in Q1 2020 (+55%), gold in 2008 (+97%), or long-vol in 2008 (+60%)), then it would still take a 10% allocation to reduce drawdowns by 10%, and a 20% allocation to reduce drawdowns by 20%.

—

My takeaway is there’s not much point allocating less than 10% to an asset class. Unless it’s able to produce Herculean returns either over an extended period of time, or reliably during a crash, then it’s not likely to add anything meaningful to long-term returns.

For the risky/equity side of your portfolio, unless your asset class is likely to return 10% or more a year (which it won’t, let’s be honest), then you’re going to need to a hold a larger chunk of if in your portfolio to have any real impact. For me, that’s a 10% minimum.

For the diversifier side of your portfolio, if your safe asset manages to rise 20%-30% when equities fall, then it’s not going to affect your portfolio’s drawdown by any meaningful amount unless you hold at least 10%.

So for me, “too small” is less than 10%.

Obviously this will vary by investor – others might consider smaller differences in long-term returns and smaller drawdown reductions meaningful. In which case, a smaller minimum allocation makes sense.

But you certainly won’t see me making a 3% allocation to gold any time soon.