wilmott.com

Hi all,I've got a set of implied volatilities here for options on energy. They are of the order 20-40%. Tio check these magnitudes, I grabbed some historical data on the same energy underlying (spot prices) and calucated the vols, to find them of the order 200%... the energy is UK natural gas. Now, my question is, why the in-consistency? I figured that the implied vols were implied from options of FUTUTRE contracts on the energy (Blacks formula for commodity options), and thus the implied vols were actually the implied vols of the futures contracts, NOT the vols of the energy as I originally intended. But then I thought, why should the vol of the futures be much smaller than the underlying energy? Surely, if the underlying is a very volatile asset, then the future contract on it would also be volatile, unless there are some major side considerations... What makes the vol of a future less than the underlying?Any energy experts out there? Regards,Sam

Sam,Two points:1. The volatility of natural gas futures contracts has a term structure such that the volatility increases as the contract moves toward expiry. This is due to the fact that spot natural gas is primarily affected by operational concerns and the immediate weather. As the weather and operational concerns are difficult (impossible) to forecast many months out, the futures contracts trade to the expected price, however, as more information becomes available in the market the price becomes more volatile. Also of note is that while the spot / futures convergence must hold, the volatility of a daily spot will almost always be much higher than the prompt month futures contract. In the US at the end of february the March contract went from 6.30 to 11.00 + in one day, but some spot markets for day ahead gas were trading at gaps of 20 dollars or more to open the day. 2. Secondly it depends upon the model you used to value the option. An option on a future is not priced at the options current volatility but rather the average volatility between now and expiry. Unfortunately Black 76 assumes constant volatility so the mistake is easy to make. There are models out there that assume stochastic volatility so it is possible to take into account the term structure of volatility. Finally, option models often give very perverted implied vols because energy prices are extremely non-normal with a drifting stochastic vol (unfortunately methods of capturing these price behaviors for purposes of valuation lead to even more rediculous implied vols). This is why I almost always use historical vols and correlations instead of implied.cheers

2 more points:When you have mean reversion, the futures vol is much lower than the spot vol. If you can't see how this intuitively, I can email you a proof.Secondly, a lot of the vol in spot prices takes into account seasonal factors and time factors (in the case of power). As spot prices will be affected by the change in season, but futures prices won't, futures prices won't incorporate this deterministic movement.

Sam, basically you answered your own question correctly: the implied vols are derived from prices of options on futures. The underlying processes for spot and futures are different. The spot is clearly subject to jumps that show in your analysis under the guise of O(200%) vols. The futures don't fluctuate as much. The reason is, crudely, that the futures contract represents a month worth of undelying. The jumps in the daily spot prices will average out if you take appropriate expectation.

Thanks to everyone on this thread!... No doubt, I have been thinking of this over the past day... and it is begining to make sense. But I am still difesting the stuff and need more time to organise it in my head gjlipman, Yes, please do forward me that proof. The best address to you use:samy@theinternet.demon.co.ukMany thanks to you all,Sam

SamUK gas options are usually traded based on the the forward product for the whole month, are you comparing day ahead or within day spot prices with options that cover gas for the whole month? Day Ahead will be very volatilile as supply and demand varies, the month less so, eg day ahead in may may hit 23p/therm say on a tight day but you wouldn't want to pay 23p/therm for the entire month on the back of that day.StoppedOut

Stoppedout,This is where I am getting confused.Hypothetical Example. Say we had 2 forward contracts, both expire on the 30th April. The contract is for 100 barrels of oil. Contract AOn the 30th April, the contract requires the transfer of 100 barrels of oil. It is delivered on the 30th April, and money changes hands (the agreed forward price when the contract was written) on the the 30 April.Contract B100 barrels of oil are delivered UNIFORMLY over the month of May. I.e. 100/31 barrels per day. Money changes hands however, on the 30 April for the whole 100 barrels. Clearly, these are both forwards, but they are different. First and foremost,1. Will the prices of these 2 contracts be the same? As Stoppedout mentioned, they can not be as in the former, the entire value of the forawrd contract depends on the price of the underlying on 30 April (if it subbenly spikes, then clearly, the forward contract has high value). The latter's value depends on each days price in May. So the prices of both contracts must be different. 2. Will both contracts exhibit the same volatility? I accept that spot vol us much higher than the vol of contract B, because of the 'averaging effect' of contract B over the month. But what about contract A? Does this mean that it will be just as volatile as the spot? or will its volatility lie between that of spot and contract B? Or am I just wasting time in even considering Forwards of type A?Kind Regards,Sam

QuoteOriginally posted by: sam. . . Or am I just wasting time in even considering Forwards of type A? . . . Yes

Thanks Tonyc,So in this case, how would one theortically get the fair price of a forward? Assuming c, u to be interest, storage costs and convenience yield respectively, and lets say interest is zero, would it look something like:F(t,T) = sum_{i=1}^{n} W(i).S(t). exp( \integral_{t}^{Ti}} (c-u) dt)where Ti are the delivery dates and W(i) are the delivery volumes for each delivery date? (e.g. For delivery over a month n is about 30). This is basically the standard arb free argument for the cost of a forward, but spread over several delivery dates. Thanks,Sam

QuoteOriginally posted by: sam. . . This is basically the standard arb free argument for the cost of a forward, but spread over several delivery dates. . . . yep, you've got it, basic cash and carry arb, but the beauty of commodities [or the complexity], is that markets are often "backwardated", and cash and carry arb goes out the window . . . of course this is just crude oil. heating oil futures, electricity options, and swing contracts in nat gas are more of what you refer to as "type a" forwards

Many Thanks for your help,Sam

Rather than start a new thread I thought I would continue here. Basically this is a questio about the settlement of futures on Crude oil. There are essentially 2 ways to settle, Exchange Futures for Pyhsical (EFP), which is largeley what this thread has been about, and the cash settlement which I am now battling with. As I understand, on the IPE trading in the futures ceases 15 days prior to the first delivery date. At this point the holder can decide between EFP or a cash settlement against the settlement price (which is the closing price on the next day). What exactly does the latter mean? What cash flow occurs on settlement? I thought that these contracts are market to market, so that any payment is incorporated into these daily margin calls... in this case there shouldnt be any cash flow at settlement. Any clarification is very much appreicated,Sam

Anybody?Thanks,Sam

I think it is slightly strange that the parties would agree to settle on a price yet unknown, as if they were agreeing to something that isn't clear...I would imagine that upon agreement to settle, they must by then also agree on the spot price to be used against which the contract is measured. The cashflow never really occurs immediately upon agreement, settlement takes a couple of days, but at least the parties know where they stand as the price is agreed...this is how the stock market settlement system works, and I can't imagine that the fundamental principle of certainty would for any reason be displaced in the commodities market. The whole point of futures, other than for speculation, is hedging...you aren't hedging anything if you agree to a value that is yet determinable!

michael, As a lawyer you will appreciate my pedantic response in saying that it is not strictly true to hedge one requires a fixed price in advance. Asian options, or asian forwards (price = average of preceding x days) do not have prices that are fixed in advance and yet they are definitely hedges. The asian element gives up some certainty with respect to price, but makes it more difficult to manipulate the derivative and underlying for a profit - which is a real worry in commodity markets, particularly energy.gammashark