From the desk of John Arnesen
Consulting lead, Pierpoint Financial Consulting
In 1990 when I entered the world of securities lending, the fee split arrangements with almost every beneficial owner were relatively straight forward and uncomplicated. An unindemnified client split the revenue with us as an agent 60/40, in favour of the client while for a client with indemnification the split was 50/50. As I worked for one of the largest agent lenders as measured by programme size, I think it's reasonable to use this as a proxy for the market in general at that time. On the face of it, this seems perfectly understandable; after all, there is something inherently 'fair 'about sharing something in equal measure between two parties. We all do it in life all the time. Dr Susanne Shultz of Manchester University in a study comparing animal and human behaviour stated,
"Unique human traits include generosity, teaching and imitation. Our model suggests the key to both of these behaviours might lie in how we overcame the impact of competition, allowing us to share resources and information between us".
In summary, she found that animals tend not to learn this when competing for highly valued foods, whereas better outcome results when resources are shared in the way that early humans learnt.
When having lunch with a colleague, you split the bill equally. Perhaps you buy your significant other a bar of Lindt chocolate? Chances are, you get half. You order 12 garlic snails as an hors d'oeuvre at Coq d'Argent in the City of London, and your dining partner gets six (actually, I begrudge that one, they are to die for). Sharing appears to work well, and when the outcome is a financial benefit to both parties, an equal fee split is a reliable and reasonable place to start.
So why have securities lending fee splits shifted so much in favour of the beneficial owner since the early 1990s? Has the concept of 'fair and reasonable' been replaced by other factors that regularly produce fee splits of 90/10 and higher ( in favour of a rarefied few investors)? Who and what is driving this?
Before I delve into that, let me address the subject of indemnification. As I mentioned, in the early 90s there was a relatively uniform consensus that indemnification ( I assume the reader understands what this is, if not, call me) was worth 10% of the fee split or 20% in additional fees. Indemnification has two elements to it; what is the capital cost associated with the risk to the agent lender in providing indemnification, and what is the risk premium the beneficial owner places on receiving it? I will hazard a guess that in the early 90s, agent lenders had a vague sense that it had a price and cost to it, but if pushed, they may have found it challenging to specify an actual figure. As a contingent liability, someone, somewhere in the organisation, would have had a more accurate sense of that cost, but the business and that individual didn't cross paths that often. The value to beneficial owners was blatantly obvious. To be able to mitigate counterparty risk in favour of the agent lender and the organisation they represented was both massive risk mitigation and the lifting of a burden in doing extensive counterparty risk analysis on an ongoing basis. To give up 10 per cent of the revenue sharing arrangement represented value to beneficial owners and was acknowledged as such.
Of course, this doesn't mean there were not more favourable arrangements. Sovereign wealth funds and central banks have always had more favourable terms, mostly based on the sheer size of the reserves which was predominately held in US government debt. Lending that asset class has always been incredibly lucrative.
Around the early 2000s, something changed.
Most noticeable was the rise in demand. The frenzy for prime brokerage services on the back of increased hedge fund demand was driving all kinds of innovative arrangements. The once hallowed ground on which only agent lenders trod, namely talking to beneficial owners was being trampled on by prime brokers (PB) seeking to secure guaranteed access to supply directly via exclusive arrangements that guaranteed an annualised revenue stream. Effectively, some beneficial owners were persuaded that PBs would pay a higher premium to have exclusive access to their assets under mutually agreed terms. The PB could draw down all, some or none of the supply at their discretion. As a result, the basis of a mutual sharing of revenue between beneficial owner and agent lender was challenged by this new alternative route to market. (To avoid getting into the weeds, I will write about exclusive arrangements in my next blog.)
Unless the agent was paid its prior fee-split out of the exclusive arrangement revenue stream ( which, by the way, was sometimes quite lucrative), it had to come to some arrangement with the beneficial owner to support the borrowing activity of the PB in terms of loans, returns, settlement costs and the consumption of resources. As most custodians did not yet provide this service within the custody world, the effort typically resided in the securities lending business and believe me when I write that it became gruesome—more in my next blog.
Once borrowers became eager to secure guaranteed supply to compete with each other for hedge fund mandates, beneficial owners were at least willing to listen to various pitches for exclusive arrangements. Seen as an existential threat by agent lenders, one tool that was deployed, albeit crudely, in my opinion, was to change the shared fee split. If the agent lender could demonstrate a similar revenue stream to that on offer by a PB, many beneficial owners were persuaded to remain within the agent's programme. Many, however, is not all, leading to several agents guaranteeing annual revenue. Now you had a situation with higher fee splits, guaranteed revenue and the provision of indemnification. Just imagine asking an asset manager to guarantee investment returns but here was securities lending ostensibly doing just that!
More demand led to more supply but to win that new supply agent lenders now had to be super- competitive. This placed a new pressure on securities lending to deliver an aggressive estimated revenue figure, especially in instances where it was a subset of a global custody bid. Of course, as with any bid, you had to agonise with wondering what the rest of your competitors would do. If you thought one of the large US lenders would offer a 75/25 fee split, you went with 80/20. Rarely did the prospective client ask about indemnified or non-indemnified pricing, but if they did, you were reluctant to make the usually quoted differential, and inevitably you offered your best and most aggressive pricing including indemnification. Then, if unsuccessful with securities lending as the key variable factor in the process, even if only implied, you felt the pressure of failure within the securities lending business. Does this sound familiar?
The Global Financial Crisis (2007-2009) resulted in several beneficial owners withdrawing from securities lending programmes. For many, it was a temporary suspension. Still, for others, it was driven by the financial loss incurred by forced selling of reinvested assets during a period of intense, negative volatility. That led to a slew of new RFPs post-crisis. The shift in demand was then centred on agent lenders that were desperate to win this ‘new’ business given that adding clients organically was an unlikely scenario for some time. Sure enough, 85/15, 90/10 and even more extreme fee splits frequently appeared.
Sadly, the industry has done this to itself. There is nothing wrong with competition, it is necessary to drive innovation and creativity, and deliver client value. However, a response solely in the form of higher fee splits does the business, and to some extent, its clients a disservice. Running a securities lending business is expensive. There are myriad costs borne by the agent lenders' share of the fee split. In assessing the value of a relationship, it begins with an assessment of the client's assets: to know the trading style of the portfolio managers, the expected volatility and turn-over via direct dialogue or as an information request in the RFP process. This detail can be hard to get, I know. The asset class and country of equities are vital in this assessment. If the portfolio is heavily weighted to Asian assets that represent 50% of the revenue estimate, the movement costs are far higher than those of a sovereign debt portfolio custodied at one of the international central securities depositories(ICSD). Other questions arise, for instance:
Has the client any specific restrictions that limit the ability to generate revenue?
Where are they located?
How frequently do they expect face-to-face review meetings?
Are there any non-standard reporting requirements that require resources to fulfil the obligation?
These are not negative assessments; they are just considerations to include in the matrix when assessing the costs. Now, imagine, for example, the same prospective client has a terrific government bond portfolio and permits term lending, and that aspect of the portfolio represents 30% of the estimated revenue with the balance made up of European equities. Would it not be reasonable to split the revenue by asset class? The cost to generate the revenue from the government debt portfolio is minuscule compare to the equity assets and can bear a higher fee split. I understand the argument that you simply blend the assessment of each into one figure and offer that, but that doesn't protect the agent's costs from a change in the overall portfolio and nor does it present the opportunity for them to demonstrate to the client it has done some extensive analysis in its approach to the fee split. I am convinced that if an agent lender is transparent with a prospective client and lays out exactly how the revenue will be generated, with which counterparties and why some are better than others, coupled with disclosure of costs it will incur in servicing the client, the chances of a mutually beneficial arrangement are far higher.
I have participated in numerous RFP post mortems, and only on rare occasions was the securities lending fee split at the top of the considerations. When it was, I wonder if that is business the agent lender wanted? Earlier in this blog, I mentioned that excessively high fee splits do clients a disservice as well. Here is why. When an agent lender is pressured by high fee splits, it has less opportunity to reinvest in its business and adopt the technology developments that are coming thick and fast. Those developments are ultimately for the benefit of beneficial owners and should, in time, reduce the agent lenders' costs. Driving expense out of business is a fixation, but it is hard to do that if your business is only scraping by.
I doubt we can return to the fee-sharing of the 90s, but it is a reasonable place to start. Is it not easier to explain a 50/50 fee split to a regulator that, by its nature is 'fair and reasonable' than have to explain why one client is at 70/30 and another at 90/10 with the 70/30 client generating higher revenue? Something needs to change.