A Blank Line in Dorado

Feldhausen, the 9th of March, 1835. The sweep log for last night records eleven nebulae in the region of Dorado, and the twelfth entry is blank - the object was suspected, then lost, then suspected again, and I closed the shutters rather than log a suspicion. That blank line is the honest part of the record.

The bulletin from the exchanges arrives weekly now, and this one concerns me directly: expiration weeks in the index options market show a 0.35% average premium, attributed to the hedging of the market makers. My own capital sits partly in such instruments, so the finding is not abstract to me. A structural regularity of that size, compounding across the calendar, would change what I can afford to leave in the hands of my trustees at Hawkhurst. It changes my household accounts. I will not pretend otherwise.

Yet the provenance troubles me, and I say so in my own interest precisely because I have an interest. Four of the five confirming series appear to draw on the same exchange feed - the confirmation is partial, not independent. The series begins in 1983, but the contract specifications changed in 1997; comparisons across that boundary require adjustment, or the premium may be an artefact of instrument change rather than a change in the phenomenon. This is the same error I guard against in the stellar photometry: the star that brightens across three catalogues may merely have been measured through a different eyepiece.

The deeper difficulty is that the phenomenon is slow in the way the precession of the equinoxes is slow - it declares itself only across many repetitions, and each repetition is contaminated by conditions that cannot be held fixed. No reliable data exists for the pre-1997 regime because the hedging mechanics themselves were different; that gap is load-bearing, and any backtest that bridges it silently invents the missing observations.

So my position is this: I will not move the trust’s money on a premium I cannot yet separate from its own measurement apparatus. The patience required is not idle waiting but active verification - reconstructing the result from independent feeds, testing whether the premium survives when the hedging flows are estimated rather than assumed.

To Mr. Glaisher at the Royal Observatory: plot the premium series against the estimated dealer gamma by week, and send me the residuals before the next expiration - the discrepancy, if there is one, will be the most important result.