1. What is a PRDC coupon, in trader English?
PRDC = Power Reverse Dual Currency note. Translation: You’re selling the investor something that pays a juicy coupon linked to USD rates, even though they live in JPY (or vice versa).
A cartoon version of the floating coupon at time \(t\) looks like:
\[ \text{Coupon}(t) \approx \text{Notional} \times \frac{\text{USDJPY}_t}{\text{USDJPY}_0} \times \big(\text{USD rate}(t) - \text{JPY rate}(t)\big). \]
Read this slowly:
• If USD strengthens vs JPY (USDJPY goes up), coupon goes up. • If USD rates are higher than JPY rates, coupon goes up. • Investor is happy because “I’m getting this high USD-linked coupon in my boring JPY world.”
But from the issuer side (the dealer who sold it), that’s like being short a mash-up of FX and rate options at the same time.
Desk intuition
Think of it like this:
You promised the investor: “If USD stays strong and USD rates stay high,
I’ll keep paying you this elevated coupon.”
That’s selling optionality on two markets at once.
2. Of course we added a callable
Because coupons could explode if USD is strong and USD yields stay rich, we (the issuers) added a call feature:
Issuer call: “If coupons get too good for you (the investor), we take the note back early.”
Everyone pitched that as “protection,” but this also stretched maturities out to 20–30y so that we could offer a high first-year coupon and still make the economics work.
What that really did:
• You sold a callable structure with very long optionality. • You inherited long-dated vega (sensitivity to long-end volatility). • You inherited cross gamma between FX and rates.
3. Two risks that actually kill you
Risk A: Long-dated vega flip
When USDJPY goes up (USD strong), coupons balloon and you can call the note back. That means effectively the optionality is getting knocked out soon. Vega exposure (to very long-dated volatility) actually falls.
When USDJPY goes down (USD weak), coupons shrink, note doesn’t get called, maturity behaves longer, and you are now stuck with huge long-dated vega.
So in bad scenarios (USD weak, safe-haven JPY bid) you suddenly become short liquidity in ultra-long vol — along with every other issuer on the street.
If everyone needs 20y+ vega on the same day, guess where implied vols go. They explode. That was one side of the pain.
Risk B: Cross gamma (a.k.a. “short correlation”)
Your P&L is not just “if FX goes here I win.” It’s “if FX goes here AND the rate curve does this, I’m still OK.”
You’re effectively betting “these two markets will move in a coordinated, friendly way,” and you’re short that bet.
4. Visualizing cross gamma with moving blocks
Let’s build a mental animation. Imagine two sliders:
When you’re running this book, you’re not just sensitive to each slider. You’re sensitive to the combination.
Cross gamma = “what happens to my P&L if both sliders move at once?”
If you’re short cross gamma, your P&L tends to be fine when markets move in your ‘normal’ pattern (e.g. USD up goes hand-in-hand with certain curve behavior), but gets wrecked when the relationship breaks.
5. Why “short correlation” is the same story
Quick analogy: take a worst-of put on two assets, A and B. If correlation drops (assets diverge), worst-of puts get more valuable for the buyer because typically one asset tanks more than the other. The seller is effectively short correlation.
Same in PRDC space, just messier: you are effectively short the “nice” correlation between FX levels and long-end vs short-end rates.
Desk cartoon of the hedge
• If USDJPY rallies (USD strong), you tend to buy long-term bonds, sell short-term bonds. • If USDJPY sells off (USD weak), you flip: sell long-term bonds, buy short-term bonds.
You’re hoping that funding spreads / basis behave in a way that offsets those flows.
But in stress, markets don’t politely offset. They reinforce.
6. The 2008-style stress scenario
What happens in a real crisis scenario like the classic stress: USD funding stress, JPY as safe haven, basis dislocates?
• USDJPY falls hard (JPY strengthens). • Long-dated callability disappears → the note “lives longer.” Now you’re stuck wearing massive long-term vega you didn’t want. • At the same time, cross-currency funding/basis moves in the “wrong” direction. The curve move doesn’t hedge your FX move — it doubles the pain.
That is short cross gamma in plain language: “Both sliders moved, but instead of cancelling, they teamed up against me.”
The lose/lose corner
• Investor: coupon collapses, call never triggers, note maturity effectively extends to 20y–30y. • Dealer: forced to hedge ultra-long vega at the worst time, and hedge rate/FX interaction in a market where liquidity is one-way. • Everyone: “Wait, wasn’t this marketed as safe carry?”
7. Why desks say “be long convexity”
Convexity is the property that big moves help you instead of kill you. In PRDC-style flow, issuers sold that convexity twice: once in FX vs rates, and again across maturity (because of the callable feature).
Being short convexity in two dimensions is what made this lethal. You weren’t just betting on where USDJPY goes or where long-end USD yields go. You were betting that the relationship between the two behaves “normally,” forever.
Spoiler: relationships don’t behave in panic.