Interesting. So you'd execute the risk premia sleeve in the futures market too? What if a flash crash happens again while it's long—wouldn't the position be at risk of being forcibly closed at a loss by ADL? Or is the idea that it should have flipped short before the market crashes far enough for ADL to be triggered?
Yes. Theoretically, an ADL type scenario could happen at any time right, either bull/bear market. And true, the model could be net long at that time. But these events tend to happen when there's an implosion of big players, a weird left tail event, etc. Those tend to happen when market is in a depressed state (see earlier similar times). This means that a trend model would generally be net short during those periods. You can even run a simple simulation up to the 10/10, a traditional trend model would've been net short at that time. So yeh, the short side is there to mitigate those events, but you can't really eliminate that risk. That's the risk you bear when you're trading a market that is not as mature as others. Also why you should seek a higher rate of return to compensate that risk.
Another thing is that when you're trading spot on an exchange, you're also bearing counter-party risk (eg. ftx). You can more those spot funds to a cold wallet and manage it there, but it adds to the infrastructure burden. So there's tradeoffs of risk in each decision.
That makes sense. Yeah absolutely, using the spot market and having to transfer to a hardware wallet would definitely increase the operational burden (albeit not too much given that a typical risk premia harvesting strategy would only rebalance once a week, at most). I suppose using a DEX is another option, but that's probably not without its risks either. As you say, you can't remove the risk entirely, and if you were able to, you shouldn't expect to earn a such a significant excess return.
Thanks for the replies. I wish you the best of luck with the new approach!
Good on you for being transparent about everything—I'm glad your portfolio lived to tell the tale! I suppose one solution would be to separate the portfolio into two: an unleveraged, long-only risk premia harvesting sleeve in the spot market (to capture the long-term positive drift in crypto); and a market-neutral alpha sleeve (to diversify and generate excess returns).
Yep, that's very similar to the solution I am going for. Not necessarily a long-only model, because you can still capture that premia when it flips "really long and really short", while having that short leg to smooth it out during bad markets and hedge against similar events.
All remaining will be alpha, net-neutral premia, etc.
Interesting. So you'd execute the risk premia sleeve in the futures market too? What if a flash crash happens again while it's long—wouldn't the position be at risk of being forcibly closed at a loss by ADL? Or is the idea that it should have flipped short before the market crashes far enough for ADL to be triggered?
Yes. Theoretically, an ADL type scenario could happen at any time right, either bull/bear market. And true, the model could be net long at that time. But these events tend to happen when there's an implosion of big players, a weird left tail event, etc. Those tend to happen when market is in a depressed state (see earlier similar times). This means that a trend model would generally be net short during those periods. You can even run a simple simulation up to the 10/10, a traditional trend model would've been net short at that time. So yeh, the short side is there to mitigate those events, but you can't really eliminate that risk. That's the risk you bear when you're trading a market that is not as mature as others. Also why you should seek a higher rate of return to compensate that risk.
Another thing is that when you're trading spot on an exchange, you're also bearing counter-party risk (eg. ftx). You can more those spot funds to a cold wallet and manage it there, but it adds to the infrastructure burden. So there's tradeoffs of risk in each decision.
That makes sense. Yeah absolutely, using the spot market and having to transfer to a hardware wallet would definitely increase the operational burden (albeit not too much given that a typical risk premia harvesting strategy would only rebalance once a week, at most). I suppose using a DEX is another option, but that's probably not without its risks either. As you say, you can't remove the risk entirely, and if you were able to, you shouldn't expect to earn a such a significant excess return.
Thanks for the replies. I wish you the best of luck with the new approach!
100%, there's really no way to remove risks, usually just shift them elsewhere.
Appreciate your comments also mate. I'll be sharing here updates on what's working and what's not.
Good on you for being transparent about everything—I'm glad your portfolio lived to tell the tale! I suppose one solution would be to separate the portfolio into two: an unleveraged, long-only risk premia harvesting sleeve in the spot market (to capture the long-term positive drift in crypto); and a market-neutral alpha sleeve (to diversify and generate excess returns).
Yep, that's very similar to the solution I am going for. Not necessarily a long-only model, because you can still capture that premia when it flips "really long and really short", while having that short leg to smooth it out during bad markets and hedge against similar events.
All remaining will be alpha, net-neutral premia, etc.