Gold slid to $3,988 an ounce on Wednesday before steadying near $4,036 by Friday morning, a retreat that has erased most of the gain the metal posted during the worst months of the Iran war. The reversal in energy has been just as visible. The iShares S&P/TSX Capped Energy Index ETF trades around $23.95, roughly 17% below the 52-week high it set while the Strait of Hormuz was still closed to shipping.
Neither move has fully registered with the investors who built these positions during the spring. That gap between what the market has already done and what an account still assumes is the disposition effect, the pattern Hersh Shefrin and Meir Statman documented in 1985: a strong tendency to hold losing positions far longer than winning ones, in the hope that a paper loss will reverse before it has to be realized.
Two Rallies, Only One Was the War
Gold's path through 2026 has actually involved two separate advances, and conflating them is part of why the current retreat feels more alarming than it should. The first rally took gold to an all-time high of $5,589 on January 28, driven by central bank buying and a broader shift away from US dollar reserves, well before the war began on February 28. That advance had nothing to do with Iran.
The second rally was the one tied to the conflict. Gold had fallen back to around $4,099 in early February, then climbed through the spring as the Strait of Hormuz closure dragged on, reaching $4,792 by mid-April and holding in the $4,450 to $4,770 range through May. That war premium, not the January peak, is the one now unwinding. At $4,036, gold sits almost exactly where it traded in early February, before the war had moved the price at all.
Why the Account Still Looks Like April
The investor who bought gold at $4,700 in May is not thinking about the January peak. The reference point that matters is the price paid, and prospect theory, the framework Daniel Kahneman and Amos Tversky introduced in 1979, explains why a loss measured against that reference point feels roughly twice as painful as an equivalent gain. The natural response is to wait rather than sell, even as the news driving the original purchase keeps fading.
Recency bias compounds the problem. Three months inside a closed Strait of Hormuz, with prices jumping on every UKMTO shipping alert, can start to feel like the normal state of the market rather than an unusual one. The Islamabad Memorandum, signed June 17 to reopen the strait, did not register immediately because the volatility it was meant to resolve had already become the baseline against which everything else was measured.
Gold's price since the start of the war shows the two advances clearly: a January peak unrelated to Iran, a second climb through the spring tied directly to the Hormuz closure, and a retreat that has now erased nearly all of the second move.
Gold's mid-April peak coincided with the height of uncertainty over the Strait of Hormuz. The dashed reference line marks the early-February price, before military action began on February 28.
The Energy Side of the Same Trade
The TSX energy sub-index tells a parallel story on a shorter clock. Suncor, Canadian Natural Resources and Cenovus, which together make up more than 60% of the iShares energy ETF's holdings, all rallied hard while WTI crude held above $90 during the closure. Friday's session saw the ETF fall nearly 4% as WTI dropped below $70 for the first time since February 27, the day before the war began.
That is the same round trip gold has made, on a faster clock. Energy positions opened in April, when the trade looked durable and the Strait of Hormuz showed no sign of reopening, are now back near where they started. The behavioural pattern is identical: a position built on a premise that has substantially reversed, held in place by the same reluctance to lock in a loss that Shefrin and Statman described four decades ago.
None of this means the premise was wrong in April. It means the premise had an expiry date, and prices have a way of moving well ahead of the accounts built to reflect them.