There has been much interest generated in the UK in recent weeks on the question of fund costs. A study showed major variation in fund charges for pension funds, often uncorrelated to performance, and this raised a question that impacts funds all over the world. The charges are all, in one form or another, a levy against a fund’s assets, and, as the recent study explained, fees are a multi-billion dollar annual business and a key component of under or over performance by funds. The question is whether such fees are justified, or even equitable.
I have always considered there to be three core types of charges and they can be defined as management, transactional and structural. Management charges include investment management and depository or fund administration fees. Transactional charges include brokerage commissions, stamp duties, custodian or clearing fees. And structural charges, the most opaque of all, include bid offer spreads, deposit rate haircuts or collateral and similar friction costs. But the fee environment is even murkier than this. There is a different structure to charging by prime brokers, especially those with an economic interest in re-hypothecation or a stake in the transactional flow of a fund. A custodian, sharing in stock lending fees, may similarly use this source of revenue to offset ad valorem or transaction costs.
I have to admit that I find it hard to agree, as some demand, that there should be industry pricing standards in the market. The funds’ business is a bespoke affair rather than an off the shelf one, unless the fund in question is a basic index tracker or tied to a popular benchmark. And, as happens in many other benchmarking services, fundamental differences in the economic relationship, as noted above, between the fund and some, or all, of its service providers is likely to distort any comparison.
It would be arrogant to put forward a solution to the problem which greater experts than me struggle with. But there are some basic rules that we can consider. After all, the global custody business has a revenue flow of around US$ 40-50 billion per annum and the global investment management community must generate revenues of several multiples of that figure. Annual charges, of that scale, are without doubt a major influencer of performance.
So what data is required to improve on transparency?
Let us start, though, by tackling performance. Lost in the paranoia about measuring against benchmarks or other engineered criteria is the fact that the logical performance measure of any fund, over time, is their performance against an inflation adjusted absolute return. That is the measure that shows if a fund is at least maintaining real value and is much more demanding than the softer, non-inflation-linked measure adopted by many absolute return vehicles. Obviously, there is relevance in a fund’s performance against its selected benchmark. Would it not help, though, if there were a regular analysis, and justification, of the delta between the performance against the benchmark and an inflation adjusted absolute return index? This analysis should logically allow for a market cycle to ensure that short term variations are eliminated. Is a three or five yearly analysis illogical? The purpose of such longer term analysis would be to allow us to rate funds by performance and thereby better assess their cost base to establish if their rankings in both correlate closely.
So what fund costs should be identified? I would suggest that figures showing the overall cost of operation may be reasonable, as long as there is consensus on common data components. I suspect we need to avoid disclosure at function level. In situations, where I have seen such transparency, the exercise often has the perverse effect of driving down visible charges and promoting more intangible sources of revenue for a manager or other supplier. I can recall, in my earlier career, fairly fabricated switches by broker fund managers whose sole purpose was to generate commission on a bad day. And some stock lending activity has definitely been more profitable for the intermediary than the legal owner.
On the cost side, we can identify certain asset related specific charges which should be expressed as a function of the assets under management. These are investment management, fiduciary, depository and non-transaction related custodian fees. Then, an estimate is needed of transaction costs. Logical costs here are brokerage fees, transaction related custody fees, stamp and other transaction taxes or levies as well as any clearing fees. These, in turn, should be expressed as a cost per average transaction.
An estimate is also needed of the more opaque fees. Here, help is needed from experts to give acceptable estimates of bid offer spreads on securities, foreign exchange, stock lending or swap transactions, the cost of collateral, especially in cases such as initial margins at CCPs, or the aggregate opportunity cost of unremunerated, or under remunerated, cash balances. These should then be allocated to the asset or transaction cost component as appropriate. There will still be challenges in assessing the opaque charges, such as how to compare “spreads” in order driven versus quote driven markets and what rate or period to use as a benchmark for cash.
But key would be the explanation as to why there is a delta. Prime brokerage is a classical example, though, of the frailties of this analysis given the cross subsidy of several of their services against the value of flow business or the accessibility of the clients’ assets for financing. But custodian banks are similarly impacted, not only where they retain an economic interest in stock lending transactions but also where advantageous fees may reflect the value of the sponsoring company’s core banking business or the flows they garner from the relevant investment management group. However, the call for transparency should also include a health check to identify, rather than necessarily condemn, such events.
Politically it is wonderful to call for transparency. However, we need to ensure that any such transparency works to improve fund performance. We also need to be careful that this exercise does not lead to degradation in service provision or the risk profile of the appropriate funds as a quid pro quo of downward pricing pressure. Attractive as the cost analysis process may appear to some, let us keep it as simple as we can. We do not want another pointless and costly data mountain that confuses rather than adds value. In the end data will show either correlation with peers, accidental correlation with peers or divergence; key is that there are experts who understand the construct of each fund and can assess the appropriateness of the data.