The devaluation of the yuan by the People’s Bank of China (PBOC) for the third consecutive day on Thursday, August 13 has had an immediate effect on the structured products market: increasing the need of local banks for liquidity, leading international banks to buy short-term options, while also increasing their long-term option positions, according to a senior structured products banker in London.
The principle effect of the devaluation was experienced most profoundly in foreign exchange target redemption forwards (Tarf), but also structured products where the underlying is also based on the fortunes of the CNY against the US dollar, says the banker.
A target redemption forward is a strip of knockout forwards. The most common use has been by Chinese exporters who want to sell dollars and buy yuan, says the banker. Banks offered them, for 24 months, a monthly settlement where they sell, say, $5m and buy renminbi at around 6.40 at a time when the spot and, importantly, the forward were trading below that level. In return for that enhanced trade, if the spot rate or fixing at expiry is above the strike, the client sells twice as much notional: this is a simple leveraged forward in FX.
The complex feature of the product is an autocall. The entire strip – from one to 24 months – is terminated and cancelled as soon as the client has made a given amount of profit, says the banker. That given amount of profit is usually between three and six months-worth of expiries when the spot does not move, ie. the PBOC holds the fix. The client has effectively sold volatility because he has sold the leveraged forward, assuming that the PBOC will keep the cap on the dollar:renminbi.
If the cap is held, the client makes his profit, and the trade will autocall after three to six months and be terminated. If the cap is not held, spot rallies a lot higher and they are trading twice as much notional at strike, which is now 20bps below the spot and forward levels.
From a bank’s point of view, the local banks have a risk they cannot manage because they do not have exotic FX books, says the banker. The local banks immediately decide to trade this product with Corporate X and then back out the exotic risk completely to an international bank. The corporate client faces the local bank, which faces the international bank. The international bank will be collateralised facing the local bank, but the local bank will not be collateralised facing the corporate, because very few corporate clients have collateral agreements, certainly in Asia when facing a small, local bank.
In this case, there is a funding and liquidity issue for the local bank: the spot has gone a lot higher, which means the trade they have backed out to the international bank is now worth a lot to the international bank because it has done a trade where it has the right to buy dollar:renminbi at 6.40 when the forward is trading at 6.80. The international bank has a big, positive mark-to-market and can go to the local bank and say ‘please can you post us $100m’. Realising it has no collateral agreement with the corporate client, the local bank has to find $100m. “One of the knock-on effects of this is it puts a strain on funding in the local market,” says the banker. It is presumed that at least some of this funding comes from the local, short-term money markets. “This shows the ramifications of the structured products market slightly outside of FX,” says the banker.
The local bank has no effect on the FX market because it has given all this away to the international bank, says the banker. What is the international bank is left with? On the trade date – the expected knockout date was after three or six months – the bank has bought a vanilla option with a tenor of three to six months. The bank is long six months volatility so will sell three and six month vol in short-dated options. To marry up the strikes, what they tend to do is sell options around the target forward. So they tend to sell 6.40 call options.
The international bank sells those options, makes money and that money pays for the upfront premium they pay to the local bank to put the trade in the book. “If spot does not move, that is absolutely fine,” says the banker. “The problem comes when spot moves. If spot goes higher, the expected knockout date of the underlying product – the target forward – drifts to the end of the trade, ie. two years (from three to six months). If spot did not move, the client accumulates enough profit for the product to autocall at the profit cap, at around three to six months. Now that spot is at or above the strike, the client is never going to make any profit, so there won’t be any autocall. So, it’s now a two-year trade.”
Now, in its options book, the international bank has, on one side, a position that looked like three to six months, but has now turned into two years. That means it has a lot longer vega (the measure of an option’s sensitivity to changes in the volatility of the underlying asset), because the option is longer. “On the hedges, where they sold those 6.40 strikes at three to six months, they are short short-dated options. And spot has now gone to the strike of those options, so the gamma (rate of change of the delta in relation to the price of the underlying) of those options has gone up, but the vega has not changed too much.
“The bank is left with calendar spread, where it is short three- to six-month options and long two-year options (because the target redemption has drifted out to two years),” says the banker. “They are getting longer vol when the vol of the market is going up, which should make them happy. The problem is, they are getting a lot shorter gamma because the gamma of a three-month option is a lot bigger than the gamma of a two-year option. The problem is the long vega position does not move as much because it is two-year vol; if you look at China two-year vol, it has moved nowhere near as much as three- and six-month vol.”
The outcome is that banks have to buy back some of the short-dated options sold to get rid of that gamma position, says the banker. That has driven the market as the banks rehedge their positions and also deal with huge spot positions: “if you think about the delta of the option: if you think about the underlying target forward when the banks put it in the book was like being a long a low delta option. Spot has suddenly flown higher, so they are now long not a high delta option (rising from around 20 to around 70), so you have those positions and also to roll out your forward hedges (to two years).”
The banks end up needing to do long-dated and much bigger hedges at the same time as the market is going a lot higher. “Trading desks have ended up being way too short short-dated options, and too long long-dated options,” says the banker. “They tend not to sell the longer-dated options because they are happy to be long vol and that is not a liquid part of the market.
“They are selling dollars forward a lot more: even though spot is rallying higher, there is a strange effect in that the long-dated, six-month and one-year swap points were going down as spot was rallying higher, which is the complete opposite of what you would expect,” says the banker. “This was because everyone rolling their hedges out: buying six-month and selling two-year, which pushes the curve down as spot rallies, which is a really weird correlation. That is all to do with rehedging coming from swap books pushing out their positions.”
The upshot is there is no interest in these products. The market was driven by corporate clients, and all of them have seen dollar:yuan fly higher and all doing twice as much notional at a strike that is 10-15bps below the spot level. The demand has evaporated after a sequence of three-month knockouts and rollovers into new deals. “No one will trade another one of these products for another 18 months, because there needs to be a flow of corporate cashflows to back the trade, and those cashflows are all tied up for two years,” says the banker.
Related stories:
China Market Review - July 2015
China’s hedging costs soar as index futures fall