Revenue Estimation from Citadel's Data Exhaust

While a lot of folks are allowing their attention to be hacked by the latest tweet, squirrel or AI-vowel combo to float by - and the entire landscape begins to suffer from an increasing left-brain imbalance - let's not forget the value of authorship. You know, the thing that reminds you that your "AI" is still just a bunch of nested if-then statements - and can't possibly solve the puzzle you are about to solve for the distant foreseeable future. That kind of "AI" is a snob that can only function on data that is somehow cleaner than reality...

Turns out, there remains a ton of valuable data lying around - as I have been saying on our Feed lately - hiding in plain sight. And, it turns out, that that data has been disclosed in small, hard-to-reach, chunks precisely because it's quite valuable once distilled properly, like a fine Kentucky bourbon. And, once distilled properly, authorship can be applied to coax that data to sing songs like Mercury or paint pictures like Dali. And, every once in a while, even the raw, seemingly innocuous data will cough up a pearl that need not be distilled at all.

In the chart below, Alphacution has built part of a storyboard around one such pearl: 2010 revenue of $888 million for Citadel Securities used to calculate that year's SIPC assessment. Now, it's not clear yet if this pearl is disclosed by error or by some force of serendipity. And, finding it doesn't necessary mean that the fog is entirely cleared away; only that the fog is diminished enough for Alphacution's contextual modeling assets to use it as crowbar to pry open another can of intelligence...

Pictures and much more of the story here..