I say none of the above, especially in light of yesterday’s story that Saudi Arabia is going to increase oil output.
The price pattern is not typical of a liquidity event, such as arises from a trading or algo error. In such an event, the price usually retraces quickly to its pre-error level. That is, price impacts are almost purely transitory if it is understood that the order was an error, or simply a demand for liquidity. Indeed, a paper by Bayuksahin, Haigh and Harris documents that prices of trades subsequently identified by the exchange as errors reverse well before the official announcement: the market participants figure it out on their own.
Here, the price fell dramatically, then only partially recovered. Thus, there was a persistent/permanent component to the price move. This combination of temporary and persistent price impacts means that market participants viewed that there was some likelihood that the order/orders was/were submitted by somebody with private information. (The fact that the move occurred on all major energy markets simultaneously supports that view, although with algorithmic trading one cannot be as confident: one can be more confident it wasn’t 4 or 5 fat human fingers.)
Put that together with yesterday’s Saudi announcement (into which they were probably dragooned by a threat from the IEA and the administration to tap the SPR) and the most likely explanation is that the announcement was leaked, or that the individual or individuals behind the decision decided to make a little money on their information and control. Which they did, because prices traded off more after the story was released publicly.
Therefore, IMO Monday’s event was an oil leak: somebody spilled the news of the Saudi decision, and that triggered a big trade that pushed down prices.
Not a tech issue. Just old fashioned front running of an announcement.