It is with much trepidation that I write about fees. It is a delicate and sensitive subject. Nevertheless, here goes.
Those of you that know me would also know that my favourite joke goes like this: two asset consultants meet for coffee, and one says “How is your wife?” to which the other replies “Relative to what?”.
So it is with investment management fees.
The only way we can assess whether a manager adds value is by comparing them to some form of benchmark. And just like two flies on a wall, or taxes, it is a national sport for service providers to try and manipulate the benchmarking discussion in their favour.
Most managers would like to compare themselves to the lowest benchmark possible so any outperformance can be portrayed as ‘alpha’ and command active management fees.
My role, as someone who allocates money to managers, is to get the best outcome. That often entails a forensic examination of a manager’s process to see if there is genuine value added (e.g. Jensen’s Alpha) or whether excess returns are something else.
Even if excess returns are not pure alpha they may still be worth paying for – just not as much as if the excess returns are genuinely uncorrelated manager skill.
This often makes me look like a grumpy old man, as I reject most of the proposals that I receive. I sometimes wonder whether grumpy people make better CIOs or whether the nature of being a CIO and having to investigate and reject so many strategies makes one grumpy. Maybe it’s like Volvos. Do bad drivers buy Volvos because of their perceived safety record, or does driving a Volvo make one a bad driver?
Let us take a look at some examples of outperformance that, in my humble opinion, are not alpha and not worthy of active management fees.
Smart beta is not alpha
Smart beta (or as I like to say, marginally less-dumb beta) is mostly characterised by strategies that can be created by an algorithm or a different index. Examples include style-based equity strategies that follow the quality, value or growth indices, a fundamental index, an equally weighted equity index or one of the many sub-sector indices (e.g. like Mag 7 or tech stocks). Most of these indices are available in ETF form and have cutesy tickers.
My proposition is quite straightforward: these strategies are clearly just a different beta and the only fee that should be paid for them is a reasonable implementation fee.
Despite being a cheap and cheerful way of getting certain exposures, these strategies have a much more important use to me: they can be used as benchmarks to assess whether a purported active strategy really adds value. For example, if a certain manager’s performance tracks the low-vol index so closely that it has a correlation of over 90 per cent, then any outperformance over the ASX200 is more likely to be beta than alpha.
Another example: if a manager’s performance in every dimension is just a small-cap value skew then any outperformance is likely to be structural beta. After all, if it walks like a duck and talks like a duck then it is a duck.
So next time you see a return series that beats the broad index ask yourself whether there is an underlying structural reason for the outperformance. If a growth-style equity manager has beaten the MSCI over the past 12 years, then that outperformance is likely to be part beta and part alpha. A quick and easy way to get an idea as to how much is alpha and how much is real value add from security selection is to compare that manager to a growth or quality-growth index; only the return above the style index is likely to be alpha.
And, if the returns of a manager track the growth index very, very closely, then it is more likely that the excess returns achieved over the broad index are almost wholly beta – suggesting that the manager is in reality a one-trick pony.
Credit is not alpha
I could scarcely believe my ears the other day when I queried a credit manager’s fees and the response I got was “Well, those fees are very cheap for the 250 bps of alpha achieved”.
When I picked my chin up from the floor, I mentioned that BBB securities carry a credit spread of at least 200 basis points and that all that the manager had done was purchased lower grade credit. Any excess over cash rates was mostly due to investing in higher-risk, lower-grade securities. This point was reluctantly conceded by the charming BDM (BDMs are mostly charming).
The only fee that should be paid for adding credit risk is an implementation fee. Active investment management fees should only be paid for performance above and beyond the additional credit risk. In this example, 2 per cent of the outperformance was likely credit beta and only 50 basis points was likely due to manager skill.
Incorrect benchmark
A well-known way to enhance a fund manager’s reputation is to use a different and lower benchmark for comparison purposes – sometimes even a different beta altogether.
Examples of this include:
- Comparing the performance of an equity strategy to cash or CPI/CPI+
- Comparing the performance of an infrastructure strategy to CPI+
- Comparing the performance of a global equity strategy inclusive of emerging markets to a developed markets benchmark (or vice versa)
- Comparing the performance of a style-based equity strategy to a broad index
- Comparing the performance of a private equity strategy to an eight per cent preferred return
- Comparing the performance of a hedge fund with some equity correlation to zero per cent
So, how much should a manager be paid? My answer is a fair fee for implementing the strategy (a base fee or account keeping fee) plus a fair share of the value added. The base fee can range from a very low number for passive traditional assets up to 50 basis points or so for more labour intensive strategies like property, infrastructure and private credit.
And what, you may well ask, should managers be paid for their skill?
Maybe investors can expect between five and 10 times their fees in alpha for long-only strategies in bonds and equities. This represents an alpha capture for the asset owner of 10-20 per cent. By way of a worked example, Australian equity managers might charge 25-30 basis points for a core strategy with expected outperformance of 2.5 per cent per annum.
In more esoteric strategies where the alpha is purer a fair fee might be three to five times the fees in expected alpha. This represents an alpha capture for the asset owner of 20-33 per cent.
In broad Australian equities mandates (e.g. ASX300) one can get the index for 2-5 bps using an ETF. So a fair total fee for core Aussie equity managers might be 5 bps plus a share of outperformance.
As an example, if one hoped to get 10 times the fees paid to managers in outperformance, a formula that might describe most flat fees for Australian equity managers would be 5 bps plus 10 per cent of expected alpha.
Let us see how well that predictive formula describes the flat fees on a $100m Aussie equity mandate
Prediction for passive 5 bps, actual 3-4 bps
Prediction for core Aussie equities (2.5 per cent alpha) 30 bps, actual 30 bps
Prediction for concentrated Aussie equities (6 per cent alpha) 65 bps, actual 65 bps
Bottom line
It costs similar to appoint a good and a bad manager, so pick good managers first and then think about negotiating fees. Just picking the cheapest managers is a poor strategy.
There is no scarcity of good managers, so just as always do your homework and think about the metrics you use to judge fees and performance.
But I have some final words of caution. Firstly, fees are just one constraint on an investment program and are just one budget an asset owner has to deal with. Others include tracking error, liquidity and regulatory risk. Fees alone should not determine manager selection.
Secondly, I prefer bang-for-buck for assessing fees rather than just net excess return. It is important that a fee budget is spent efficiently as well as wisely. There are lots of very expensive managers that prefer the latter. That’s a subject for another article.
Finally, if it is not investable, it is not a good benchmark. CPI + 4 per cent or an 8 per cent preferred return might be a reasonable long-term expectation but is useless as a short-term performance measure for longer-term strategies. And we are all long-term investors aren’t we? (He says rhetorically, coming from an industry where managers have a shelf-life of three years maximum before underperformance causes too much pain).
But in these difficult times, one thing you can control is fees – so make sure you are getting value for money. Every little bit helps.
Michael Block is an adjunct industry professor at the University of Technology Sydney, and has held a number of chief investment officer and allocation roles over his 40-year career in investment management.
