Having watched with interest the various talking heads waffling up a storm regarding the ICIS Heren market manipulation scandal it is patently apparent that almost all of them don’t really seem to have much of a clue as to what went on, what could it have affected, and why someone would have tried to do it in the first place – always assuming that attempted market manipulation is proved to have occurred.
So for those insomnia afflicted readers who would like to gain some insight – please find herewith my observations on the issue.
In the price gathering window between 4:30pm and 5:00pm on the date in question, when ICIS Heren reporters rang around their various tame traders and brokers asking what prices they had made on their last deals; a number of trades were reported that – upon analysis – seemed to be below expected market numbers (i.e. the rest of the ring around). These numbers are used to determine the settlement price for the Heren Day Ahead index for gas. This index is used to settle the price paid for day ahead gas on many contracts and is also used to calculate strike prices for options written against the index. It is therefore a key price in the derivatives market and a key price for day ahead trading which is effectively the spot market for UK Natural Gas. Physical contracts deliver into the UK system on this price where a producer is selling gas to a counterparty on the day ahead index rather than on a forward price or at the within day system price. Big North Sea producers who have chosen not to hedge their gas production often sell gas at the Heren Day Ahead price.
What in effect is alleged is that some reported trades done almost on the close (4:30pm for settlement price purposes) were struck at a price below the prevailing price indicated by the rest of the market. Is there an innocent explanation for this? Possibly – if a trader found himself too long (owning too much gas) at the end of the day he could have priced it to sell – hence the low price offered and hit (bought). Is this likely? Without knowing who the counterparties to the transaction were and examining their positions at the time it is difficult to make that call – hence the report to OFGEM and the FSA. It is cause for investigation possibly; but with any investigation into malfeasance the first question should be asked “Cui bono” – or who profits. In this case it is very difficult to make a case for manipulation although I am not aware of the volumes involved in the trades. Let’s outline the possible villains.
Who dun it?
- A short (seller) of a massive call option (the right to buy gas at the strike price) with a strike price at or marginally above the level of the market that day who did not want to have to pay out against that call they get to pocket the premium for the option (which also represents their underlying position as being short the settlement price). The issue with this theory is that options in the gas market settle against the average of the months Heren Day Ahead index. Therefore a manipulation that alters the settlement price on a single day from, say, 55p/therm to 54.8p/therm, has only affected the strike price of the call by 1/30th of 0.2p/therm – or 0.007p/therm. In this case a 10million therms per day call (underlying market value of gas at 55p/therm being £165million for the month) would net the manipulator a whopping £21,000 assuming that the strike price would have been breached without the manipulation which is by no means certain. I think we would need to look for a particularly idiotic trader who wanted to risk jail for a marked to market PnL improvement of £21,000 on his book. We would also need to be looking for a trader that writes derivatives – not thick on the ground in physical trading houses (suppliers and producers) and these guys are many things but idiotic about money is rarely one of them!
- A long (buyer) of a massive put option (the right to sell gas at the strike price) with a strike price similarly positioned to the above example. This is effectively the opposite position to the first hypothesis – but the motivation is the same – they get to sell gas at a price higher than the settlement price (which represents their underlying position as being long the settlement price). The rest of the maths follows. The natural buyers of put options are producers – so your big oil boys. They might have the volumes mentioned in the previous example but this would be a massive risk for minimal gain – same as the derivative traders risks.
- A physical player who has sold gas on the OTC day ahead market and has a contract to buy gas at the Heren DA settlement price. Here we are talking trading the underlying itself – so 10million therms for delivery that day would benefit from the full 0.2p/therm differential created on the day. However that still only nets £20,000 (excluding the loss made on the gas sold below market to achieve the manipulation in the first place). We cannot discount however that the trader was not looking for a bigger impact than 0.2p/therm so maybe they were shooting for a 1p impact rather than the 0.2p achieved. To achieve that level of move would require a bucket load of optimism or rigging on a scale that couldn’t fail to be noticed – as we can see even this size discrepancy was spotted.
- A physical player who has bought gas at the Heren DA settlement price and has sold it within day on the OTC markets. Here we are talking about the reverse position of the above scenario – but all other maths and motivations apply.
- Lastly we can look at the possibility of a very sophisticated derivatives seller who has a digital put option paying out at the strike price mentioned in scenarios 1 and 2. Here a 0.2p/therm move might be enough to alter the settlement price enough to avoid the digital strike. A digital option takes an option payout from being one of an incremental difference – in this case 0.2p/therm, to being one with a larger digital payout – say 3p/therm. Now you are moving into real money territory. The thing with a digital option is that it still settles on the monthly average, is an extremely complicated derivative and thus not only require a sophisticated desk to write it, but also a sophisticated physical producer able to both understand it, value it and obtain permission to buy it. To say these circumstances are improbable is to understate how unlikely this scenario is.
The reality of this market manipulation story is that in order to make money at it as a derivatives trader you need to manipulate quite a few days in the month – not just one. Both them and the physical trader need the move to be bigger or to keep risking their job by multiple manipulations. Or else they need to manipulate something else – the month ahead settlement price for example. Conversely they could simply go to the power market and float thick, off the market power prices at the back end of the curve and make a tidy sum from market illiquidity rather than from manipulation. I know which one I would be trying for – simply based on the risk reward dynamic. Of course there could be a devious digital trader at work here, working in tandem with the shooter on the grassy knoll, and pounding on an unsuspecting oil major. This would be a shame.
Did this affect you and I and our central heating costs? No. Does it suggest widespread market manipulation by the wholesale market traders? No. Does it highlight, by virtue of an absence in recent (or any) suspect price reporting in the electricity a problem with the gas market? No. As FSA investigator Sherlock Holmes might say to his Watson; “The curious thing Watson, is that the electricity dog never seems to bark at all. Liquidity my dear Watson, liquidity”