Multi-Prime 3.0

Like any useful product or service, the use of multiple prime brokers by hedge funds has evolved over the last fifteen years. We are now entering our third major iteration of the multi-prime model. Before we look at version 3.0 let’s take a brief tour of the earlier versions of the multi-prime model.

Version 1.0. Pre-Lehman Brothers. The first era stretches from the late 1990’s to September 2008. One of the key enablers was the advent of hedge fund administrators vertically integrating and offering middle office accounting and support services to hedge funds. This removed a critical obstacle to hedge funds having more than one prime broker. Hedge funds no longer needed the staff and accounting systems to manage more than one prime broker. They could now just connect to their administrator and bolt on as many primes as they wanted or needed. Hedge funds start adding primes for reasons that include funding rate transparency and broadening stock loan access.  Larger funds add primes to spread their wallet and to get more services like research and corporate access. Some small and medium (and even some large) size funds still have only one prime broker.

Version 2.0. Post Lehman Brothers bankruptcy. The next version kicks in September 2008. There is nothing like a global catastrophe for everyone to get religion fast.  The multi-prime model morphs quickly to become all about mitigating counterparty risk.  Institutional investors start demanding it and single prime is effectively dead. PB diversification is still only within the top ten bulge-bracket banks (“big ten”)[1].  Larger, more sophisticated funds start diversifying by bank type and region.

These more sophisticated hedge funds even start probing their prime broker’s sub-custodian network in an attempt to root out exposure to banks to which they are trying to avoid (think Citigroup in November 2008).

Custodian banks are quickly added to the mix as hedge funds begin shipping in cash and unencumbered securities from their prime brokers. The rub here is that the large custodians have little or no experience dealing with hedge funds on a daily basis like the prime brokers, and the services provided are generally perceived to be sub-standard.

On the derivatives front hedge funds accelerate requests for tri-party arrangements to hold initial margin at large custodians. Due to resource constraints, at both investment banks and custodians, only the largest hedge funds are able to negotiate such arrangements.

Version 3.0. Post-Basel III. This is where we are today. Concerns about insolvency risk fade with the new capital, liquidity and leverage rules and are replaced by uncertainty over the cost of financing and the availability of bank balance sheet.

Prime brokers, feeling pressure from their treasury and funding groups, start evaluating their hedge fund clients based on return (usually on assets i.e. balance sheet) in addition to revenues. Hedge funds that are below the ROA hurdle rate are asked to re-structure their financing balances or face price increases or possible expulsion from the platform. Hedge funds, some now using as many as eight to ten primes, are buffeted by daily calls from their primes asking them to shift positions and financing. The prime brokers, in a brilliant stroke of marketing disingenuousness, start advising their hedge fund clients to reduce their number of primes in order to boost their ROA at their remaining providers.

Requests for consolidating primes makes sense if a hedge fund client’s ROA is above the hurdle rate. Then the additional financing balances result in increased revenue, and more importantly, improve the overall ROA of the prime broker’s book. However what is the incentive for a hedge fund to concentrate more of their financing (and equity) when their ROA is already above the hurdle rate?[2] So primes must want their clients that are below the ROA hurdle to consolidate providers and financing. But this doesn’t make sense either. If these hedge funds increase their overall level of financing with a PB, without altering the mix of these balances, then the impact will be to lower the overall ROA of the book while leaving the client’s ROA unchanged. The prime brokers do not want this. What the primes really mean when they suggest that their clients reduce their number of providers is to give them financing balances that improve the client’s ROA i.e., a financing mix that is complementary to the client’s current book.

The prime broker consolidation argument has another critical flaw. It only works on a smaller scale because prime brokers have limited balance sheet. Would the primes feel the same way about consolidation if a large number of their clients decided to move in significant financing balances (like 2008)? Would the offer still stand?  I don’t think so.

So, do hedge funds need more or fewer primes? I’ve already addressed this in a recent post[3] and my view is unchanged. More primes. Balance sheet is scarce and hedge funds using leverage are going to get stopped out at their primes. They will need additional providers to finance their portfolios.

The issue in Version 3.0 becomes one of portability. Hedge funds, realizing that bank balance sheets have become a scarce resource, are starting to have concerns about the portability of their book between primes.  How much additional financing can their current prime brokers absorb?  The top ten banks all start to look the same though the lens of Basel III, i.e., limited, and contracting, balance sheets. The portability of the hedge fund’s financing book, a given in the pre-Basel III world and the foundation of the risk mitigation benefits of the multi-prime model, is no longer certain. Version 3.0 continues to unfold. Stay tuned.

[1] Morgan Stanley, Goldman Sachs, JP Morgan, Citigroup, BAML, Deutsche Bank, Credit Suisse, UBS, Barclays and BNP Parisbas.

[2] Unless of course the hedge fund wants to increase revenues to get access to additional services e.g., corporate access.

[3] “Safety In (Prime) Numbers”

Leave a Reply

Your email address will not be published. Required fields are marked *