Article: What risk disclosures tell us about Goldman’s problems and Morgan Stanley’s success08 September 2017 | Nick Dunbar Feed
2017 has been difficult for those at Goldman Sachs who are accustomed to being at the forefront of US trading firms. Half-year revenues at Goldman’s fixed income, currencies and commodities division declined 20% year-on-year, accelerating to a 40% decline in the second quarter.
Goldman’s new chief financial officer, Martin Chavez, added to the sense of disquiet when he admitted on a July earnings call that “we didn’t navigate the markets as well as we want to and as well as we typically have”.
Contrast that with Goldman’s old rival Morgan Stanley, whose fixed income trading revenues surged in the first half of the year, up 36% from a year earlier. The bank’s success appears to be more than a fluke: its return on RWAs has been the highest for banks in the Risky Finance database for over a year now.
This ratio is more important than the absolute level of revenues because it shows how much revenue the bank makes per unit of risk it takes. If a bank can boost this ratio and keep it high, then it can control revenues by adjusting risk levels.
To get high RoRWAs, banks first have to achieve a one-off reduction in risk. Our time evolution chart shows how Goldman and Morgan Stanley started in 2012 with very high market RWA numbers, then brought them down as Dodd-Frank and Volcker Rule reforms began to bite.
Screenshot of an interactive visualisation available to Risky Finance subscribers.
At the end of 2014, Morgan Stanley started reducing market risk faster than Goldman, a move that coincided with a 1,500 cut in trader headcount. By 2016, Morgan Stanley’s market risk appetite was closer to universal banks such as Deutsche than its traditional downtown rival. Some observers thought the firm wouldn’t be able to recover from such belt-tightening.
That view turned out to be wrong. As Risky Finance data show, in summer 2016 Morgan Stanley got into the high RoRWA zone. Then it dialled up its risk appetite again this year, pulling ahead of Goldman in trading revenues.
Goldman looked on the verge of entering the zone last year, then fizzled, with RoRWAs falling almost as low as Wells Fargo this year. The firm stayed flat in its risk appetite this year while revenues tanked.
There are a number of theories about Goldman’s stumble. The firm’s commitment to commodity trading is said to have resulted in losses this year, pulling down FICC revenues. Some investors worry that the firm has cut back trading staff too aggressively – but then again the same charge was levelled against Morgan Stanley.
These theories also ignore the fact that Goldman looks much more like a commercial bank than Morgan Stanley does. Its credit RWAs as a percentage of equity have increased since 2012, and now are higher than JP Morgan or Citigroup, and double that of Morgan Stanley. Yet if Goldman looks like a commercial bank, it doesn’t think like one, and hasn’t achieved the high RoRWA transformation that JP Morgan or HSBC have done.
Perhaps the divergence between Goldman and Morgan Stanley is something to do with their respective risk cultures. Goldman’s chief risk officer, Craig Broderick, was closely involved in his firm’s controversial subprime hedging decisions during the crisis. He became widely known when he later defended the firm’s actions to Congress.
Morgan Stanley’s CRO, Keishi Hotsuki, has a lower profile. He made a well-timed exit from Merrill Lynch in 2007 then headed Morgan Stanley’s market risk department through the peak of the crisis, ending up in the CRO seat in 2011.
Perhaps Hotsuki’s lack of financial crisis baggage makes him more nimble in steering his firm’s risk positioning today. Consider how Morgan Stanley’s market risk metric which has risen the most this year – stress VaR – is precisely the one which replicates financial crisis conditions on today’s portfolio.
Meanwhile, Broderick hasn’t been able to control Goldman’s credit RWAs and as a result hasn’t got the freedom to crank up market risk like Hotsuki can. We don’t know much more because Broderick doesn’t talk to analysts or give interviews. Goldman is currently in the process of separating Broderick’s department into a standalone unit with a greater focus on information security and operational risk.
It remains to be seen whether this move will solve the problem, or if the firm needs a completely new approach to capital management, and new people to do it.