Process for Returning FTX Funds Announced

There are two mechanisms likely at play here -- fraudulent conveyance under 11 USC 548 and applicable non-bankruptcy law doesn't have a 90 day limit.

I don't see how F/P provided new value to the insolvent FTX parties, or any value at all, in exchange for the transfers. The theory behind new value is that the other creditors are just as well off as they were before and so there has been no preference.

Replying to my own comment, I suspect that the next step will be demand letters from the FTX bankruptcy estate to grantees. Those may come sooner than I had expected; I was somewhat surprised to see so much attention given to charitable/political donations so early in the process, given that the bulk of them are fairly small in the grand scheme of things. That could be due to the media attention given to these donations.

A few observations on how that could happen -- of course, I cannot predict or offer advice to any individual grantee, including an assessment of any potential defenses the grantee might have. If the Madoff matter is any guide, the estate may be willing to negotiate at that point. The Madoff trustee's stated goal was "to negotiate, not litigate, and [his procedures were] designed to facilitate the evaluation and administration of each case and encourage an out-of-court, amicable resolution." The Madoff trustee also dismissed, or decided not to bring, 533 avoidance actions on the basis of demonstrated hardship (although I believe that was a novel program). 

There was apparently some consideration of the "whether former customers used any fictitious profits to pay taxes" in assessing hardship claims (p. 40 of document below). I infer that the trustee was willing to consider that money spent to pay taxes on fictitious profits generated no benefit to the potential defendant (to the extent the potential defendant couldn't benefit from taking a tax loss in the year the loss was discovered). In contrast, the trustee doesn't generally seem to have been concerned about potential targets having spent money to their benefit that they wouldn't have spent but for receipt of fictitious profits.

You can read more about how clawbacks went down in the Madoff matter here.  One outcome of note was that the trustee settled one major claim (Hadassah) for about 58% of the alleged avoidable amounts to prevent closure of a non-profit hospital. Also, there was $84 million from a charitable foundation that "the Trustee did not seek to recover, because the funds had been donated to charitable causes" (p.  73); it is unclear whether this decision was made for legal reasons or because recovery from downstream charities who did not have any dealings with Madoff seemed inequitable. In general, the "benchmark for good faith claims at 95 percent, in consideration of the comparative ease in making such cases" (p. 77), but the discussion is focused on larger targets and I would guess that there was more flexibility for smaller ones since the costs of litigation do not scale linearly with amount demanded.