Centralisation: are we putting all of our operational eggs in one risky basket?
The starting point
In February 2018, a 23-year-old trader posted a message on Reddit to WallStreetBets, a forum dedicated to trading. The message was just four words: ‘Should I kill myself?’
The trader, username Lilkanna, did not kill himself and remains active on the site. But he learned a painful lesson about putting all of his eggs in one basket: the prompt for his despair was losing $4m going all-in on an inverse volatility contract, XIV.
Along with sympathy, the most common response seemed to be amazement that anyone would risk everything on a single bet – even the most casual retail trader knows the importance of diversification, right?
But perhaps the rest of us aren’t such perfect risk managers ourselves. Perhaps we’ve been putting our eggs in one basket this whole time?
Centralisation and operational risk
At ITRS, our focus is more operational risk – and its converse, operational resilience – rather than trading risk. But this domain has its own eggs-in-one-basket challenges that are cause for concern.
Centralisation of key infrastructure is the culprit here. If a single bank or broker fails, then so be it. The business will suffer and its clients will be inconvenienced but ultimately, they will find somewhere else to trade and life will move on.
But central institutions such as clearing houses and exchanges? That’s a different matter, as became clear in February when CME Group had an outage and some of the world’s biggest markets were untradeable for three hours. It was lucky that many major markets were out of hours and others saw limited price moves, meaning traders didn’t lose out by as much as they could have.
And it’s not just the rarefied world of professional finance that can be affected by centralisation. In retail, roughly 95 per cent of payments depend on Visa. So when a 10 hour outage hit the platform last year, it affected 5.2 million payments across Europe, imperilling everything from weekly shops to business acquisitions.
It’s a trend we see time and again, in anything from financial services to gaming networks. It begs the question: is centralisation a mistake?
Resilience, not diversification
No, centralisation is not a mistake. It is a strategy that supports technical efficiency, cost-savings and economies of scale. No one is arguing for the breakup of CME Group or Visa. Done right, centralisation can be a powerful tool for innovation and effective risk management – it can be a driver for, rather than a threat to, operational resilience.
The key is to enshrine some key operational resilience best practices to prevent outages. Resilient architecture – including failover wherever appropriate – is an obvious starting point. Any bottleneck or single point of failure should be mapped and monitored with robust recovery plans.
Capacity planning and load testing are also crucial: stress testing just how much systems can withstand before failing and then carefully monitoring for when systems are approaching capacity limits. Internal resources can then be redirected or additional capacity procured to head off the threat of outages.
So, are we putting all of our operational eggs in one risky basket with the trend for centralisation? Perhaps in some respects, but instead of spreading the risk, which will require greater oversight and cost input, we should be focusing on fortifying the basket we do have with a move towards relentless and meticulous operational resilience.
