Gold rose. Why did the call option lose value?

A correct direction is an incomplete explanation

A trader buys a gold call, gold rises, and the option is worth less at the next observation. That sequence is possible without any contradiction. The trader owned a particular option with a strike, expiry and premium, not a promise to receive every upward dollar movement in bullion. The review must therefore begin with the purchased instrument rather than the correctness of the directional headline.

CME’s vega lesson identifies implied volatility as a separate influence on an option premium. Its theta lesson identifies sensitivity to elapsed time and warns readers to check the model’s time convention. Those definitions provide a framework for diagnosis; they do not tell us the current volatility, premium or fair value of any gold option.

Source notes: [1] CME Group — Options Vega: The Greeks · [2] CME Group — Options Theta: The Greeks

An intentionally simplified premium worksheet

Consider a hypothetical call quoted at 30 premium points. Assume its initial delta is 0.40, its vega is 1.50 premium points per one percentage-point change in implied volatility, and its theta is minus 0.50 point per calendar day. These are invented inputs for a local sensitivity exercise, not an observed option chain or a claim that every model would produce this combination.

Now assume the underlying future rises by 10 points over two days while implied volatility falls by four percentage points. A first-order worksheet attributes plus 4 points to direction, minus 1 to elapsed time and minus 6 to volatility. The combined estimate is minus 3 points, taking the premium from 30 to approximately 27. Direction helped; the other assumed effects more than offset it. Contract multipliers and trading costs are deliberately excluded.

The approximation is a question to test

The worksheet is not a substitute for repricing. CME’s combined-Greeks lesson stresses that the sensitivities move together as market inputs change. Extending the starting delta, theta and vega unchanged through a large move can produce a poor estimate. The four-percentage-point volatility move above is useful for demonstrating competing effects, but a full model calculation would be required to assess the approximation’s size and error.

I would retain both results: the initial sensitivity explanation and a recalculation using the new inputs. Their difference is informative. It can reveal nonlinear effects, inconsistent conventions or a stale quotation. Calling the entire discrepancy “volatility crush” without checking these alternatives would replace a measurement problem with a convenient label.

Source notes: [3] CME Group — Options Premium and the Greeks

An event ending does not guarantee cheaper uncertainty

Before an economic release, I would record the option’s own expiry and strike rather than attach one volatility number to all gold options. Then I would compare the same contract after the release. If implied volatility falls, that is a measured change in the pricing input; the claim that the event caused it remains an interpretation requiring a carefully defined window.

This article has no verified current gold option chain and makes no claim that volatility fell today. A release can leave uncertainty elevated or introduce a new question. The practical lesson is to test several combinations in advance: gold up with volatility down, gold up with volatility unchanged, and gold down with volatility up. The scenario labels describe possibilities, not assigned probabilities.

Use executable sides when evaluating an exit

A model value, last trade and available bid answer different questions. For a long option, a displayed midpoint is not evidence that the position could be sold there in the desired quantity. A research worksheet should retain quotation time, bid, ask, size and contract identifier alongside the model inputs. Otherwise a stale last trade can be mistaken for a change in the option’s economics.

I would also separate the reason for owning the option from the reason for closing it. If the intended information event has passed, continuing to hold because the original directional view still feels correct creates a new time horizon. That revised decision deserves its own assumptions. It should not silently inherit the first trade’s justification.

Review the purchased exposure, not only the prediction

A useful journal asks which input changed most, whether the original scenario sheet included that combination, and whether the exit used a real executable quotation. It should preserve losing examples and occasions when the expected move arrived too slowly. Those records show whether the instrument matched the intended question.

No option return, broker fill or CY66 performance is established here. The numerical example is a deliberately simplified sensitivity illustration, not P&L evidence. The separate Daily Gold Brief reviews market information; this lesson examines why translating a directional view into an option can introduce additional decisions about time, volatility and the price paid for both.