Frankie vs Goliath: Will Dettori Day change how the industry manages risk?

Fond memories or nightmares, depending on your view of Frankie Dettori’s magnificent 7 in 1996, came flooding back on Ascot Gold Cup day as Frankie rode the first 4 winners. This sent the odds of his 5^th mount, Turgenev, spiralling south from 16/1 to 3/1 (6/4 with one major online bookmaker), before being collared late on and saving the industry £50m to £200m, depending on what estimates you believe.

Fair play to @jbchips cashing out his Frankie Dettori acca prior to Turgenev finishing second in the Britannia pic.twitter.com/Wj5rowUhTe

— ODDSbible Racing (@ODDSbibleRacing) June 20, 2019

More importantly, it saved bookmakers staring down the barrel of perhaps £1 billion+ in pay-outs were Frankie’s 6^th mount, Questionare, to also win. But as it turned out, this runner finished well down the field.

Crisis averted, for now.

Over the final two days of the meeting, at least two big firms, Sky Bet and Bet365, stopped taking 4 fold and upwards bets on Frankie rides, a decision which divided opinions, to say the least.

The fall-out was curious and raised multi-faceted questions.

One of the immediate questions after Frankie’s 4^th winner on Gold Cup day was whether cash-out is broken. This question was a consequence of the price crash on Turgenev in race 5, meaning cash-out offers were largely based on the hugely shortened odds of Frankie’s last two rides. Any punter sitting on an acca rolling into Frankie on race 5 and/or race 6 gleefully took what was in some cases a life changing offer for 4 winners. But by doing so, these punters reduced the relevant operator’s risk by many millions of pounds. cash-out, rather than being broken, was in fact the operators’ saviour.

In fact, perhaps cash-out represented the only significant hedging option open to operators who wanted to limit their black swan downside, and the cash-out offers they bid, exceeded the four-fold odds, multiple times.

Were firms writing bets they could not pay? We don’t know, but based on some estimates, at a minimum they seemed to have written bets that could have wiped out an entire quarter’s or year’s P&L. That would have sent shockwaves through share prices and racing’s Levy (which is based on operators’ profits). I would think that operators write bets of similar magnitude on most weekends’ football coupons and less frequently on racing.

These may not have come in yet but interestingly, in December 2014 and again in January 2015, punter-friendly football results (which generally means a series of big team ‘bankers’ winning) are believed to have cost UK-facing bookmakers close to £100m. If a couple of more bankers had also won during those weekends that could have indeed been 10-fold more. So yes there’s some evidence that operators are accepting very financially challenging, if not potentially fatal, bets. As noted by GVC racing trading director James Knight last week, the nature of accumulators means liabilities can spiral out of control very quickly.

Are firms obliged to write bets above their risk tolerance or at all? The crowds yell yes! However, one would assume the firms’ management would say resoundingly no. Given the industry limits winning punters all the time, it was probably prudent to limit or suspend ‘Frankie accas’ on Friday and Saturday under these unprecedented circumstances. This may well lead to a more systematic cap on accas to be seen a handful of times a year.

And perversely, it may have been the biggest firms at the most risk, as they presumably service the largest percentage of recreational punters latching onto this bet on Friday and Saturday – and bigger firms also generally have the biggest acca risk limits per client. This represents a potentially dangerous combination for them regardless of their financial firepower.

Were risk systems adequate in real-time? Probably not, but I am sure these are all currently being investigated and ‘fixed’, though what that means to each firm internally may be quite different.

Finally, it was asked in some quarters whether big operators could insure these events with Lloyds or similar. The reality is that this depends on the amount you need to hedge, as the first £10m can be hedged with on course bookies and exchanges (granted at lower prices than the operator has generally laid initially). Offering cash-out (and bidding up the price) may reduce the operator’s risk by a further 50%, if you can entice half your client base to cash-out. For online operators, this is surely achievable, though more challenging for shops. However, if you had £500m at risk, you could still be sitting on a £240m+ totally unhedged liability and this cannot be negotiated with Lloyds in 15 minutes on a Thursday afternoon.

It is possible that an insurance pool could be built, backed by global insurers and priced by professionals, but the uncertainty of the pricing and infrequent use of such a system would in practical terms prove it unworkable. Insurance firms are happy to insure a pub chain for £40m loss of revenue if England don’t reach the last 16 of a World Cup, but the reason is clear and the price is very stable and determinable. What price is the Turgenev and Questionare double with a City insurance firm for £200m of risk? The double was around 200-1 in the morning and circa 20-1 as Turgenev jumped. That is not the type of eventuality on which you make a £200m insurance contract, in my experience.

So what happens? Not much.

Operators will recheck systems and perhaps record and manage risk more ‘live’. cash-out remains the most obvious avenue for bookies to reduce risk in the heat of the battle – an option which was not available to them in 1996 when Frankie rode his Magnificent 7, though that was pre-internet with the damage largely confined to shops.

Having laid Turgenev at 5.4, it was ‘fun’ to sit back and watch the general panic and potential carnage unfold on Thursday but the industry will hopefully have to wait another two decades before history almost repeats itself.

Bernard Marantelli is CEO of Colossus Bets