Hedge only the fuel nobody else is paying for

When I joined Saltbush in March 2021, the treasury policy I inherited was tidy and well documented. It required us to hedge between 60 and 80 per cent of forecast jet fuel consumption over the following twelve months, tapering further out, using swaps and options priced against the Singapore jet kerosene benchmark. The board had approved it, the auditors had tested the hedge accounting and the analysts who followed us liked the stability it produced. On paper, the company was well protected against the commodity that is usually an airline's largest cost after its people.

The difficulty was that more than half of the fuel we burned was not ours to worry about. About 58 per cent of our flying is contract charter for resources companies, carrying their workforces to and from mine sites. Those contracts, like most in our part of the industry, pass fuel through to the customer. The charter rate has a fixed component for the aircraft, the crew and maintenance, and a fuel component that resets each month against a published index. When fuel rises, the customer pays more. When it falls, the customer pays less. We had hedged that fuel anyway.

A hedge on fuel the customer already buys

Consider what that combination does. If the price rises, the charter customer pays the increase through the pass-through clause, and our hedges on the same litres pay out as well. We gain twice on one movement. If the price falls, the customer pays less, which leaves us where we were, and our hedges on those litres lose money. We lose on a movement that should never have touched us. Taken together, a hedge on pass-through fuel is not protection. It is a position that profits when oil rises and loses when it falls, held by a company whose shareholders never asked it to take a view on oil.

Nobody had designed it that way. The policy was written when the charter business was smaller and its contracts reset fuel less often. As charter grew, the share of passed-through fuel grew with it, and nobody went back to subtract it from the hedging base. The policy measured consumption because consumption is easy to forecast. It did not measure exposure, because exposure can only be found by reading the contracts.

In September 2021 I recommended that we hedge only the fuel used on scheduled services, where we set fares ourselves and cannot pass a price rise through overnight, and that we leave charter fuel unhedged except for one narrow risk I describe below. The notional value of our hedge book fell from about A$240 million to about A$95 million. Our chief executive was uneasy. He valued the smooth fuel line in our results and expected analysts to read a smaller hedge book as less protection. The chair of our Audit and Risk Committee asked me to model the outcome under a sharp rise in prices, which was exactly the right request.

Five months later prices did rise sharply. Jet fuel climbed steeply after Russia invaded Ukraine in February 2022, and by our calculation the old policy would have produced about A$31 million more in hedge gains during the 2022 financial year than the new one did. One analyst wrote that we had left the money on the table. The board asked me to explain, and I used the same arithmetic: those gains would have arrived on top of higher charter revenue that the pass-through clauses had already delivered, and the same position would have cost us money in a falling market. In the 2023 financial year prices fell back, and the old policy would have lost about A$24 million. I take no pleasure in having been right about the second year. The point is that we were no longer betting on either.

The lag is the real exposure

Taking the hedge off charter fuel exposed a smaller risk that the old policy had hidden. Our pass-through clauses reset monthly, using the previous month's average index price. In a steadily moving market that lag is trivial. In the weeks after February 2022 it was not. We bought fuel at the new price while charging customers on the old one, and the gap cost us about A$6 million in a single quarter.

That was my error, and it taught me something useful about hedging in general. The exposure worth protecting is usually narrower and stranger than the one a policy describes. In our case it was a one-month timing difference on charter fuel, the full price risk on scheduled fuel, and the currency risk on both, because fuel is bought in US dollars and our fares and charter rates are set in A$. We now hedge each of those separately. Short-dated swaps sized to one month of charter consumption cover the lag. Scheduled fuel is hedged on a declining ladder over twelve months. Currency is managed as its own programme, since an airline can lose as much on the exchange rate in a bad year as on the commodity itself.

The narrower book brought benefits I did not fully expect. Cash posted as collateral, which had reached A$18 million when prices collapsed in 2020, has been nil since early 2022. Hedge accounting under AASB 9 became simpler to document and audit, because each instrument now relates to an identifiable forecast purchase that we genuinely bear. And treasury discussions with the board became shorter, because the risk we were managing could be described in two sentences.

What the risk committee should be shown

I am not arguing that airlines should stop hedging. A carrier that sells tickets weeks or months ahead at fixed fares has a real exposure between the day it sells a seat and the day it flies, and hedging that exposure is prudent. My argument is narrower. A hedging policy should begin from exposure, not consumption, and exposure lives in the customer contracts as much as in the fuel forecast.

For directors that suggests four questions. What share of fuel burn is already passed through to customers, and on what reset terms? Does the hedging base exclude that share, and if not, who decided to hold the resulting position? What does the lag between buying fuel and recovering its cost amount to in a month of sharp movement? And is currency hedged with the same discipline as the commodity, or left to whatever the treasury desk thinks the dollar will do?

A management team that can answer those questions has a hedge programme. One that can only report the percentage of consumption hedged may be taking a view on oil on its shareholders' behalf without anyone having decided to. A smooth fuel line in the accounts is not the purpose of the exercise. The purpose is to leave the company exposed only to the risks it is paid to carry, and to show the board, in plain numbers, which ones those are.