Last week, I focused on how to think about credit risk, relying on the bedrock of modelling – probability of default (PD) and loss given default (LGD) to show why one new rule exacerbates economic inequality. This week, I’ll turn to how this model applies to the political-risk analytics essential to effective management of the profound risks and opportunities evident across the U.S. political landscape. Adopting this analytical construct, it quickly becomes clear why companies on the sidelines awaiting clear signs of legislative action or a pronouncement from the President are engaging in the type of “risk management” that led all too many companies to hold far too many subprime mortgages in 2007. Waiting for someone to tell you risk is about to be realized is a very poor substitute for seeing it coming. It’s not the PD risk that matters; it’s the fat tail of LGD that squashes you every time.
Since the election, FedFin has provided clients with forecasts of the far-reaching changes in financial policy likely to be pursued by the Trump Administration. Think about these as our forecasts of probability of occurrence (PO). Starting in November, we outlined the Glass-Steagall 2.0 proposal that has taken considerable shape ever since. Even so, several of you have pushed back, arguing that change of such magnitude is unlikely in a country as gridlocked as the U.S. We’ve heard the same on our forecast about the demise of OLA and the pending risks to foreign banks. To be sure, clients questioning our PO calls are even-handed – we’ve also gotten some pushback when we’ve pointed to likely reforms they want – a return to state-based insurance rules, regional-BHC tailoring and stress-test relief for example.
I don’t in any way blame anyone for discounting the ability of the U.S. to do much about any financial policy of substance. But, enough has happened since the election in both politics and markets to warrant remapping the political-risk stress scenarios.
When doing so, I recommend using two risk factors – not only the probability of occurrence (PO) referenced above, but also and importantly the impact given occurrence (IGO). Let me turn quickly to each in general terms – conducting the PO/IGO analysis for each pending policy change requires more space than a simple memo affords.
First to probability of occurrence. On the plus side are factors such as GOP control of the Congress and White House, clear evidence of numerous Dodd-Frank backfires, changing regulatory personnel, and sluggish economic growth at a time when many incumbents fear for their lives. PO negatives include Senate rules, Trump Administration dysfunction, regulatory-agency staff intransigence, and the ability to stop administrative action in the courts. Weighing each of these PO factors on an issue-by-issue basis sets the stage not only for considering what might be done to you, but also whether you should pursue those policies you want done for your own benefit or the disadvantage of competitors.
With PO in hand, let’s turn to impact given occurrence. To show how this works, I’ll return to the Glass-Steagall question. Using PO to dissect what is meant beneath the Glass-Steagall 2.0 label to differentiate specific policies is necessary to a precise political-risk reading, but IGO in general terms is still clear. Almost regardless of the specifics of a Glass-Steagall 2.0 in the U.S., winners and losers will be big. This isn’t a policy that nibbles at the edges of product pricing or access to secondary markets – it’s a franchise redefiner that has huge impact across the financial-services industry. A recent FedFin report lays this out in more depth.
As a result, even if you think PO is low, IGO almost surely isn’t. To sit back and wait for the regulation and legislation laying out what will happen or, more imminent, the rules that more easily make it so is to take a huge bet that whatever happens either won’t be all that bad or won’t ever come to pass. Think about the AAA ratings on MBS – where huge losses were due to reliance on conventional wisdom, not disciplined PD/LGD analytics – to see why effective political-risk analytics shouldn’t be based on the buzz about town.