Since the 2008 crisis, marginal lending to large companies in Europe has been structurally unprofitable due, in large part, to higher capital requirements under Basel III and banks’ cost of funding broadly remaining higher than companies.
This point alone has forced many banks to either rein in or cease growing their large corporate loan books, but if that is not forceful enough then Deutsche Bank argues the cross-selling opportunity is futile too.
Bank analyst Matt Spick and strategist Nick Burns, at Deutsche Bank, have found that the amount of additional profit needed to justify a loan from businesses such as M&A advisory, FX and cash management, is implausibly large, implying there is simply not enough cross-sold revenue to make these loans into “go do loans”.
In a report on the ‘profitability of corporate lending' published on February 17, Spick and Burns broke this down, using M&A first as an example.
They started with a cost of equity for a European corporate/SME banking business at about 11%, as a “rough and ready measure”. From there they calculated how much additional profit from an alternative business a bank would have to earn to make an 11% RoE from a single-A rated corporate, on a €100 million loan. They then converted 11% net of tax hurdle rate to a 16% pre-tax required return – and 30% implied tax rate.
By estimating a 75% cost-income ratio on investment banking revenues cross-sold – and assuming the incremental revenues have no capital requirement of their own –Deutsche calculated incremental investment banking revenues required of €3.7 million for an AA-rated bank, €4.3 million for an A-rated bank, and €5.5 million for a BBB-rated bank.
 |
| Source: Deutsche Bank |
The question then is, are these cross-sell numbers plausible?
According to Deutsche, an AA-rated bank generating €3.7 million of additional revenue would be able to turn a loan to an A-rated counterparty from loss-making to cost of capital returns.
It adds that in M&A, the 10 largest acquiring companies paid an average of $90 million to the Street across an average of eight deals, or $11 million (€8 million per deal) globally. These are some of the largest transactors, and each €100 million loan would have to pick up a full fee from each transaction.
In practice, however, these fees are shared across multiple investment banks for the roles they play.
“In other words, there is effectively no way for a bank to generate enough fees on a single transaction to justify a €100 million loan and earn cost of capital,” says Spick and Burns. “The bank has to recycle the credit into either the secondary loan market or into the bond market. The loan does not belong on the bank balance sheet.”
The same is broadly true for other cross-sell products such as G10 rates and FX, or cash management. Where cash management relationships exist, they “cannot be a case for marginal lending”, says Deutsche, adding that “if we look at sales and trading cross-sell, we find it hard to see cross-sell making sense in most cases” too.
Spick and Burns argue that equities sales and trading products “would not make a natural fit” for cross-sell against loans, while FICC sales and trading products such as FX and G10 rates are capital-intensive businesses on their own, “that we believe at the industry level barely achieve their cost of capital for the majority of banks”.
Furthermore, they argue FX and G10 rates are businesses that are transitioning to electronic platforms, such as swap execution facilities, and that exchange-type swaps business will “only make it harder for banks to bundle swaps in with loans and extract surplus rent”.
As a result, Spick and Burns highlight two broader problems with cross-sell as a concept.
First, cross-selling is not the same as “coincidence-selling”. If a large investment bank sells an FX swap and an interest-rate swap to the same client, at market-competitive rates, then no cross-sell has happened. Each product has been sold on its own merits at a market competitive price, and there has been no excess profit for the bank.
Second, cross-sell can create a “misallocation of resources”. A bank might make a loss-making loan to win an M&A mandate. M&A bankers will often ask other businesses to subsidize transactions to win a deal. Adding the “M&A revenues to the loan might make the loan look profitable pre-cost allocation, but in reality it is not”.
Spick and Burns say M&A revenues and the costs need to be added, and conclude that to the extent banks, especially universal banks, continue to “press the case for cross-selling, investors need to push back on how well banks cross-allocate costs and capital”.