> FX swap markets, where for example a Dutch pension fund or Japanese insurer borrows dollars and lends euro or yen in the “spot leg” before later repaying them, have a history of problems.
This sounds speculative / risky. Is it? If so, why would a pension fund be involved with something like that in the first place? Is it typical to have a portion of a pension fund allocated to highly speculative risk?
A Dutch pension fund will have all its liabilities (future payments to pensioners) in Euros, but they typically invest a lot of the current pension contributions in USD stocks and bonds. This exposes them to exchange rate risk (uncertainty about the USD-EUR rate in the future, when they need to pay pensioners in Euros but their investments are in US dollars), on top of the normal investment risk of USD stocks and bonds. By simultaneously entering a EUR-USD swap (where they agree now with the counterparty at what rate they'll exchange USD back to EUR in the future), they can hedge the exchange rate part of the risk.
This is at least how FX swaps are supposed to be used: to reduce exchange rate risk. It could also be used for pure speculation by entering the swap in opposite direction (so the USD profits or losses get amplified in Euro terms). We don't know which. It might be in the 90-page BIS report, but (without reading) I'm guessing that the BIS researchers didn't manage to find out either.
It's primarily a matter of reducing risk by hedging. Pension/insurance companies typically hold assets around the world but have liabilities concentrated where their customers are -- and thus denominated in their customers' primary currency.
Not necessarily - it certainly could be very risky, but sometimes these types of things actually reduce overall risk, if you happen to have an offsetting, inversely-correlated risk elsewhere.
Like if you bet me $1M on a coin coming up heads, that could appear very risky on it's own. But if you already had a bet in place for $1M that it comes up tails, then overall the heads bet zeros out your risk.
A slightly better example - if the Australian Future Fund owns $50 mill of Coke bonds, they probably need an fx swap so that every time Coke sends them some USD, they can swap it into AUD and know in advance how many AUD they are going to get.
If the USD got stronger against the AUD, the swap the Future Fund are in becomes a liability for them, if viewed through a "mark to market" lens. But the USD they are receiving is worth more to them, so they could increase the value of the bond holding - this would mean everything was "on balance sheet". But there is a very good argument that this "on balance sheet" method is a bad representation of the true economics of the situation. So instead, they leave the value of the Coke bond on their balance sheet at the same level, and don't record the liability of the fx swap.
Clearly the use of the FX derivative reduces the volatility for the Future Fund.
Note this is a somewhat contrived example, but it illustrates why off balance sheet isn't a black hole per se. At the same time, if you think through a few issues that could come up in this case, it's easy to see why it can be manipulated by bad actors, and why counter party risk is a big deal.
It’s more: through various operation you end up with a large undesired $1M bet on head as a byproduct. To counter your exposure you put another $1M bet on tail. Now whatever happens, it shouldn’t impact you. Of course managing the two bets have a cost but that’s one you are ready to pay to escape being affected by the coin flip.
Hedge significantly reduces the exposure of companies to events they don’t want to be exposed to. Of course it could blow but that’s still better than having companies carrying random risks they don’t want.
So you went in with $2M (two bets on same event at $1M each) came out with $1M because only one outcome can manifest, and you call that zeroed out risk?
Or are you talking using the same million to make both bets to seperate individuals? In which case, if your tails better turns out to be full of shit and can't produce the million you were intending to pay to the guy who you owe for your bad call, you're still out your million.
> Is it typical to have a portion of a pension fund allocated to highly speculative risk?
It certainly was in the US during the 2006 housing bubble crash, for pensions and especially for city/state pension funds. The customers of pension funds and the taxpayers of cities have little to no access or say about what the funds are investing in, so extremely large funds can be sold bad investments by influencing a small number of people who are subjected to very little oversight, and plan to be gone (probably to work for the people they helped) before the investments blow up.
They are doing it because yields are good, and as a fund manager you are rewarded quarterly/yearly for good returns. If it blows up after a 5 year run, you’ve had 4.5 solid years of bonuses and walk away claiming ‘no one could have foreseen the <eminently foreseeable event>’
Look into the eurodollar system. Most "money" today is just swaps on bank ledger entries backed by counterparty trust and can-kicking. If you thought the dollar was a scam by not being backed by anything, wait till you check this out.
This sounds speculative / risky. Is it? If so, why would a pension fund be involved with something like that in the first place? Is it typical to have a portion of a pension fund allocated to highly speculative risk?