Scenario: Fed uses QE and ECB doesn't.Then CDX outperforms iTraxx and EUR appreciates v. USD.Bullish -> short CDX protection in EURBearish -> long iTraxx protection in EURIs it? If I'm bearish I don't buy protection on CDX to avoid being wrong way, correct?I know, I'm a rookie.

Makes sense to me.. Assume QE will make credit rally you'll see CDX going tighter vs Itraxx when you get QE in the US and not in europe, apart from that eur will appreciate because the fed is printing money.short CDX protection is correct / you can do this in EUR to make sure your relative notional won't decrease and your carry will keep it's valuelong Itraxx protection is corect / if you put this against the CDX trade you should do it in EUR too

Judging from your scenarios you described you (mainly) have two vanilla trades to choose from (or a combination of both):1) Sell protection on CDX (in USD), buy protection on iTraxx (in EUR) and dynamically rebalance your CDX notional with the EUR/USD exchange rate2) Buy EUR against USD as a speculative tradeLet's ignore the second option and focus on the credit (and correlation) strategyIf CDX tightens and iTraxx widens by same %PV in each currency and EUR/USD goes up at the same time, you will experience a loss. Here, the positive USD PV on CDX will be less than the negative EUR PV on iTraxx (those PVs would be equal if FX rate was unchanged). Generally speaking: the higher the correlation is in reality, the worse your trade performance will be (compared to same strategy, but lower correlation)Let's assume someone is willing to show your price in CDX quoted in EUR (quanto). Back of the envelope it should be something like this: If the implied correlation in the quanto is less than the correlation you expect in reality = You should trade the quanto If the implied correlation in the quanto is higher than the correlation you are expecting = You should trade the vanilla strategy If the implied correlation is way too high (compared to realistic expectations) might even consider separately buying protection in CDX in EUR and selling protection on CDX in USD and rebalancing the notional, hoping for a lower correlation than implied.(In reality it's obviously a bit more complicated, as we are assuming that all the potentially wild 2-dimensional scenarios are accurately captured in one single correlation parameter. That is an oversimplification at best)

Thanks a lot to the both of you, this sounds pretty good. But I have some other questions:i) Why on earth should CDX and iTraxx change by the same %PV in the respective ccy?ii) I presume the only possible estimate for the credit/FX correlation is historical. Do you think it is reliable or maybe there's something better?Please be patient with the idiot.

(i) I presented this particular example to highlight specifically the correlation effect, without having any net credit effect. Same example could be made with any PV changes (i.e. they do not have to be equal), as long as the value of USD changes in the same direction as the spread on CDX. (ii) That's the beauty of all markets. You know what has happened in the past but you never know the future. While historical data is the obvious starting point, I think it's even more important to have a strong probabilistic opinion on all the different macro scenarios that could pan out and how they affect credit, rates and currency collectively.

And you are wise.Thanks a lot!