Menu Close

UDP Provisions are Proving Ineffective Under New LNG Market Dynamics (Part I)

Pakistan’s media is abuzz with warnings of another winter gas crisis. A significant contributor to the looming gas shortage is the decision of several suppliers to cut back on LNG cargo deliveries under existing medium and long-term contracts with Pakistan LNG Limited (PLL).

Speculation is rife that these cargo cancellations are economically driven. With slopes of between 11.6% and 12%, the sales prices for November cargos under these contracts are about $8.85/mmbtu to $9.15/mmbtu, based on a Brent 301 index settlement of $76.23/bbl. In comparison, average spot prices for November delivery into South Asia exceeded $30/mmbtu in recent weeks. Redirecting cargos to the spot market does involve a cost, however. Pakistan’s contracts contain under-delivery penalty (UDP) provisions that will reportedly require the suppliers to pay 30% of the contract price to PLL for any undelivered volumes.

UDP provisions, when included in an LNG deal, are intended to penalize supplier non-performance. They should therefore disincentive attempts to optimize contracted sales positions vis-à-vis spot market opportunities. They also mitigate the costs customers may need to bear should they find themselves needing to backfill undelivered volumes via the spot and short-term markets. The problem is that UDP provisions have not been very effective of late in either discouraging the redirection of contracted cargos into the spot market, or in compensating customers for under-deliveries, whatever the reason.

We can best understand the problem by approaching this provision as a put option held by the supplier. When the spread of the spot price, Pspot, over the contract price, Pcontract, exceeds the penalty, (UDP%) x Pcontract, the supplier is better off selling the cargo into the spot market. In formula terms,

If Pspot – Pcontract > (UDP%) x Pcontract, the supplier realizes a positive payoff by exercising this option.

If we subtract the penalty from both sides, and multiply by the delivered LNG quantity, Qdelivered, we derive the following payoff formula for this option:

Payoff = Qdelivered x Max [Pspot – (1 + UDP%) x Pcontract , 0]

Assuming PLL’s UDP of 30%, and a slope of 12% against a 301 Brent index, Brent301, the payoff formula becomes:

Payoff = Qdelivered x Max [Pspot – 130% x (12% x Brent301), 0]

Given current market conditions, the option is deep in the money, even after the 30% penalty. Assuming again a Brent301 settlement of $76.23/bbl, and a conservative recent LNG spot price of $30/bbl, the payoff is over $18/mmbtu:

Payoff = Qdelivered x Max [$30 – 130% x (12% x $76.23), 0] = Qdelivered x $18.11/mmbtu

Given an average cargo size into Pakistan of about 150,000 m3 LNG, or 3.5 TBtus, this yields an incremental P&L of $63 million per cargo:

Payoff = Qdelivered x $18.11/mmbtu = 3,500,000 mmbtus x $18.11/mmbtu = $63.4 million

Clearly, the UDP provision as it was structured is not an effective economic disincentive. To understand why, we need to look at the two primary reasons why LNG contract versus spot price spreads have been behaving the way that they have.

The first is the fundamental decoupling between the crude oil and LNG markets, that has led to a collapse in price return correlations between Brent and LNG. As shown in Figure 1, the correlation between Brent crude oil and LNG price returns is now negligible at below 10%. This is a level lower than that seen between Brent and European carbon emissions allowances, ARA coal or even Henry Hub natural gas. The second cause is the unprecedented surge in LNG price volatility, which recently reached 128% .

Figure 1. Prompt-Month Price Return Correlations (as of 29 October 2021)

Source: CapraView

Both of these trends, lower correlations and higher volatilities, act to boost the value of delivery and offtake options embedded within Brent-indexed LNG SPAs. Next week, in Part II of this post, we will quantify the value of the UDP-based optionality, and highlight its hypersensitivity to changes in correlation and volatility. We will also propose remedies to improve the effectiveness of UDP provisions.