When a parallel shift assumption fails an ALM committee

Many internal risk packs still lead with a plus-or-minus 100 basis point parallel shift. That number is easy to communicate. It is also a poor description of how gilt and euro sovereign curves have behaved in recent years.

What committees actually ask

ALM chairs usually want to know which bucket hurts first when the belly sells off or the long end rallies on overseas demand. Parallel shocks flatten that conversation into a single duration number.

In our Duration Stress Review we build three paths from the client’s own holdings extract:

• A bear flattener led by the front end
• A bull steepener if long-end demand returns
• A butterfly that stresses the 5–10 year sector while wings hold

None of these require a full risk-system rebuild. They do require honest modified-duration figures by bucket and a willingness to discuss limits, not only headline DV01.

A modest caveat

Scenario packs are still stories about tomorrow’s curve. They are not forecasts. Their job is to force a clearer conversation before the next volatile issuance week.