Regular readers may know that I’m often vexed by the cost of investment funds. But there’s one element that doesn’t distress me that much. And that is a performance fee.
I suspect I’m out of kilter with many private investors on this. And the current trend in the industry certainly seems to be away from performance fees.
But I’m looking to beat a global equity index fund, albeit in a fairly limited and hopefully low-risk fashion, and I’m happy to pay a little bit more to achieve this.
On the flip side, when performance is less inspiring, I like the blow to be cushioned (admittedly usually in a very small way) by a lower fee.
How common are performance fees?
Data is tough to find but there’s an interesting note from Kepler from earlier this year that suggests some 37% of 338 non-VCT investment companies have a performance fee.
I must admit that’s quite a bit higher than I would have expected.
They are much more common with alternative funds (45%) — particularly private equity and property — than those invested in listed equities (33%).
However, the same note suggests around 51% of trusts had a performance fee at some point in their history, so they are less prevalent than they used to be.
The AIC’s annual reviews support this, saying that 6 companies abolished their performance fee in 2016, 4 in 2017 and 9 in 2018.
What’s more, a Citywire article from 2015 says that more than 30 companies had ditched their performance fee in the previous four years.
These figures may not match up precisely but the direction of travel is pretty clear.
Performance fees are largely confined to investment trusts it would seem. Kepler also has data on open-ended funds, saying that only 3.4% of the 1,321 equity-based funds they examined had performance fees.
Given that most exchange-traded funds track an index of some sort, I would have thought they are very few ETFs with performance fees.
Why the decline?
The general trend at the moment is for lower fees, driven by the popularity of low-cost passive investing. That’s great to see of course.
As well as performance fees disappearing, we are also seeing lower percentage fees and the introduction of tiered percentage fees once the funds manage to reach a certain size.
I’m certainly in favour of tiered fees, as investment management typically has enormous economies of scale. In most cases, though, the tier reductions appear pretty minimal, so I suspect the benefits of a larger fund still flow largely to the managers rather than its investors.
There is also a drive for simpler fee structures and this is one area where performance fees certainly struggle.
Here are just some of the things to consider when assessing the suitability of a performance fee:
- the appropriateness of a benchmark used;
- whether there is additional hurdle over the benchmark before a performance fee kicks in;
- the timescale it is measured over;
- whether it is based on net asset value or share price returns;
- any clawback of previously paid performance fees should returns slip back;
- if there is any penalty for underperformance (in the US, I believe, such fulcrum fees as they are called are the only type of performance fees that are permitted);
- in the case of illiquid investments, whether it is based on realised profits from asset sales or just management valuations;
- any cap on the overall charges; and
- whether there is a previous high watermark that has to be passed before any payment is triggered.
What you might think is a fairly simple exercise — constructing a sensible performance bonus — can get messy when you apply it to real-world situations.
More information required
Another irksome quality of performance fees is that you often need to dig around to get the finer details.
Ideally, whenever looking at a trust with a performance fee, I’d like to see a 10-year record with the base and performance fee split out and the returns for each year versus the benchmark. That should give a quick sense of whether the performance fee looks reasonable.
I’d also like to see more information on the AIC site regarding performance fees. The basic company profiles cover them and the statistics section shows ongoing cost both with and without performance fees, but the latter is only helpful in years when a performance fee is due.
Adding a simple yes/no token to show where performance fees exist, the percentage level and the benchmark used would be helpful extra data points, especially when you want to assess a particular performance charge against others in the sector.
What makes a ‘good’ performance fee?
In the spirit of Charlie Munger, we could use inversion here and consider what makes a bad performance fee.
Certainly, the classic hedge fund model of 2% of assets plus 20% of outperformance seems way too steep (5 and 44, though? No problem!)
Tetragon, which I’ve written about a few times, uses a tweaked version of this at 1.5% plus 25% of excess returns. The 25% is on returns over US dollar 3-month LIBOR plus 2.65%, which works out at a hurdle rate of just 4.5% right now.
This also looks way too rich for me.
But you can perhaps go too far in the other direction. The ill-fated Woodford Patient Capital Trust famously had a very basic fee to cover administration expenses with the entire remaining fee coming from 15% of any excess returns over 10% per annum.
Previously, I thought this was a good thing. But it may have contributed to some of the dubiously high valuations used for its most controversial investments and the general Hail Mary approach it seemed to take.
Performance fees in practice: room for improvement?
Certainly, I would want to see a reduction in the base fee when there is a performance fee in place when compared to other similar investment companies.
The Kepler note suggests this generally does happen with global trusts with performance fees averaging 0.65% versus 0.75% for those without. However, this is a pretty small sample size and could be skewed by the size of the trusts in question.
I couldn’t find any evidence on what the average profit percentage figure is. Anecdotally, I would say 15% is pretty standard from most trusts I have looked at recently. I’d favour 10%.
Also, from what I can tell, most performance fees seem to be measured on an annual basis, whereas I’d like to see them based on longer periods. You could look back at five-year periods every 12 months.
Lastly, pretty much all performance fees I have come across seem to be asymmetric, i.e. they only reward outperformance and don’t punish underperformance. Let’s share the pain a little.
Is there evidence they work?
There’s nothing compelling that I could find that suggests such fees work, but then it’s tough to isolate what impact they have from everything else that affect a fund’s performance.
The Kepler note looks at the UK Smaller Companies, European Smaller Companies and Global sectors. The results are mixed — trusts with performance fees have done better when it came to smaller companies but worse in the case of global mandates.
There’s an interesting study by Cass Business School in 2014 that advocates a performance-based structure but on a symmetric basis. It highlights the risk of ‘asset gathering’, where managers have more incentive to increase size regardless of the underlying performance.
I have to admit this is one thing that makes me wary of investing too much in alternative asset funds, many of which seem hell-bent on expanding as quickly as possible with significant and frequent placings.
Having a performance-fee element in such funds provides a little extra assurance that any new assets bought aren’t dragging down the overall quality of the portfolio.
The Cass study also examines whether performance fees affect the risks taken by managers, especially when they are near to high watermarks that might trigger another payment.
I hadn’t considered this specific point myself before, but I typically favour performance fee structures where the likely payment remains fairly small compared to the basic fee (a maximum of 25% perhaps).
Limiting the performance element in this way won’t eliminate such risks, but I think it should reduce them.
I’ve cooled ever so slightly on performance fees while researching this piece.
But I still think they have a place and having a sensibly sized incentive that aligns investors and managers should do more good than harm.
We want the best managers running investment trusts and a little kicker from performance fees should make the industry more attractive.
Performance fees seem more suitable to trusts run by big managers, where the individuals running the fund are less likely to own a significant personal stake.
Where funds are self-managed and the manager has a sizable holding they would seem much less appropriate. Physically holding shares should provide a much purer incentive than a complicated performance fee. I don’t think you need both.
I’m also more in favour of them with alternative funds, particularly where there’s a risk of asset hoovering for the sake of it (looking at you renewable energy infrastructure).
Let me know if you’ve any thoughts about performance fees in the comments section below. Are you drawn to them, totally indifferent, or can’t wait for the day they are abolished?
Please note that I may own some of the investments mentioned above -- you can see my current holdings on my portfolio page.
Nothing on this website should be regarded as a buy or sell recommendation as I'm just a random person writing a blog in his spare time and I am not authorised to give financial advice. Always do your own research and seek financial advice if necessary!
A free share with Freetrade
Do you fancy a free share worth between £3 and £200 plus commission-free trading? Then why not sign up to Freetrade using my referral link?
Your free share is selected at random and usually awarded on the following Wednesday once you have funded your account with at least £1. I also receive a random free share for each completed sign-up. You could get an individual company share, an ETF or, if you're really lucky, an investment trust!
Subscribe to IT Investor
Get an email alert every time I publish a new article. Your email address won't be used for anything else.