Yeah, this narrative has been around for a while and there have always been two major problems with it:
- If Bitfinex was buying Bitcoins on their own exchange with fake USD in order to push the price up, they wouldn't do so by printing Tether because USD on Bitfinex is not backed by Tether - it's supposed to be backed directly by USD in Bitfinex-owned bank accounts, with Tether only being minted as necessary to handle Tether withdrawals. (Someone even found externally-visible evidence that there wasn't enough Tether in existence to cover the Bitfinex USD balances - naturally, this was seen as further proof it was a scam.)
- During the period in question, it was Coinbase/GDAX which was pushing the price up, and that exchange never accepted Tether. The price of Bitcoin on Bitfinex and the other exchanges was pretty much always lower than GDAX, or to put this another way, Tether USD was worth more than its face value during the time when Bitfinex was supposedly printing it to push the Bitcoin price up. This is the exact opposite of what should have happened if they were doing so.
It looks, at a quick glance, like this still has the same problems.
Edit: this is more clever than the previous arguments, in that it also uses evidence from round-number trading biases and odd end-of-the-month behaviour of the kind you'd expect if Bitfinex had to make the Tether shortfall disappear for their monthly audit, and also examines outflows of Tether from Bitfinex to other exchanges. (The round-number stuff is less convincing since the divergence between exchanges and even between currencies on the same exchange was particularly bad around round numbers.)
They also attempted to test the hypothesis that the Tether printing was a reaction to demand by comparing the amount of printing with the Tether-USD price. Unfortunately, they used the price on the Kraken exchange which had almost no volume or market depth compared to the potential Bitfinex-Bitcoin-GDAX route. Basically, no-one actually used it, and so it didn't accurately represent how much it'd cost to actually exchange any substantial amount of Tether to USD or vice-versa. It's entirely unsurprising that it wasn't correlated with anything of note.
- If Bitfinex was buying Bitcoins on their own exchange with fake USD in order to push the price up, they wouldn't do so by printing Tether because USD on Bitfinex is not backed by Tether - it's supposed to be backed directly by USD in Bitfinex-owned bank accounts, with Tether only being minted as necessary to handle Tether withdrawals. (Someone even found externally-visible evidence that there wasn't enough Tether in existence to cover the Bitfinex USD balances - naturally, this was seen as further proof it was a scam.)
- During the period in question, it was Coinbase/GDAX which was pushing the price up, and that exchange never accepted Tether. The price of Bitcoin on Bitfinex and the other exchanges was pretty much always lower than GDAX, or to put this another way, Tether USD was worth more than its face value during the time when Bitfinex was supposedly printing it to push the Bitcoin price up. This is the exact opposite of what should have happened if they were doing so.
It looks, at a quick glance, like this still has the same problems.
Edit: this is more clever than the previous arguments, in that it also uses evidence from round-number trading biases and odd end-of-the-month behaviour of the kind you'd expect if Bitfinex had to make the Tether shortfall disappear for their monthly audit, and also examines outflows of Tether from Bitfinex to other exchanges. (The round-number stuff is less convincing since the divergence between exchanges and even between currencies on the same exchange was particularly bad around round numbers.)
They also attempted to test the hypothesis that the Tether printing was a reaction to demand by comparing the amount of printing with the Tether-USD price. Unfortunately, they used the price on the Kraken exchange which had almost no volume or market depth compared to the potential Bitfinex-Bitcoin-GDAX route. Basically, no-one actually used it, and so it didn't accurately represent how much it'd cost to actually exchange any substantial amount of Tether to USD or vice-versa. It's entirely unsurprising that it wasn't correlated with anything of note.