some dozen or so years ago my firm’s then head of trading was a gentleman we will call Dave. Dave had an excellent grasp of markets, options pricing, risk management and liquidity premiums but he sucked at managing people. Nice guy, knew everyone’s kids names, never yelled much but he was terrible at understanding how incentives drove behaviour.
As such, Dave took it upon himself to create an incentive plan for our group’s IT staff that was almost entirely weighted on performance time. By Dave’s thinking almost nothing was worse than wear housing risk and being unable to see it, or trade out of it due to systems issues. So the 700 pound gorilla within the IT bonus calculation was the uptime metrics on quotes, analytics and market access. Be up, get paid. Basically the Ron Jeremy plan but with more hair.
My first inclination that the plan may not work as advertised came when I asked our group IT head – who we will call Steven – for some tweaks to my what if analyzer, particularly around net gamma exposures. Stevan’s first answer was no. Assuming he had misheard my request i restated it in smaller words, speaking more slowly. Steven’s eyes smiled as he said “um no” And offered up some bullshit excuse around resources, library conflicts or network dependencies.
I waited a day and asked a third time, only to see Steven’s answer change from “no” to “fuck no”. As time passed this became the knee jerk response to any request.
“Steven can we get….”
“Fuck no, that’s impossible”
“But I didn’t even get to say what I want….”
See Steven recognized that change created opportunity for outages, which put his bonus at risk he was invented to prevent improvement.
The lesson learned is that when one pays someone to make sure a system never breaks, they also must pay them to keep the system relavent or risk having a stale dated system that doesn’t meet the requirements of the day. This is such a long understood notion there is a Latin phrase for it “SECus Phukus Marketus”.
We have created stand alone regulators whose sole benchmark is whether the system breaks. Nobody measures the success of FSOC based on asset manager AUM, or the SEC based on asset turnover, or the CFTC on open interest, or Finra on investor participation. Instead we ask them to burden business with reams of new regulation to ensure against another 2008, flash crash or cyber event.
I’m not saying we shouldn’t protect against such events, but we need our regulation to be proportional. The volcker rule alone – which is net good for my fund – is totally mad. The underlying notion made some sense, but in an effort to ensure nothing ever goes wrong we have made it so unworkable nothing can ever go right.
Now the real issue here is regulators pandering to ill informed politicians who are likewise incented to make sure the system never has a hiccup. So they allocate SEC and CFTC budgets only if they guarantee no outages. Steven is effectively running the regulators.
Instead of a pro business approach that thoughtfully marries systemic risk manaent with the need for robust functioning markets, we have an asymmetrical system hell bent on zero outage times. And I have no fucking clue what my gamma exposure will be if the market has a 4 percent move.
I know that saying the inmates should have some say in how the asylum is run comes across as asinine. But allowing disinterested parties to waterlog our markets with over the top regulation is a fools paradise.
We need to devise real, market based metrics that are used to measure and manage our regulators. Such metrics should be a happy balance of risk management and market liquidity measurements constructed with input from both regulators and practitioners. Volcker done right, with thought around its impact on the high yield and cohorts debt market liquidity should provide a safer financial system that still works. Volcker in its current form makes the market safer the same way not letting your kid ever leave the house makes her safer. The cost of the added safety is waaaay too high.
After all two condoms are safer than one, but at that point why bother.