Forgot your password?
typodupeerror

Comment: Mansfield Cumming (Score 1) 41

by Paul Rose (#36999952) Attached to: The History of Invisible Ink
Couldn't have asked for a better name...

From the last link:

Mansfield Cumming, ... was "making enquiries for invisible inks at the London University".

In October he noted that he "heard from C that the best invisible ink is semen", which did not react to the main methods of detection. Furthermore it had the advantage of being readily available

Comment: Re:Algorithmic trading? (Score 2, Interesting) 299

I'll grant you your distinction between "fake" and "false".

The orders (and news) were indeed "real".

If the intent was not to move the price without seeking a fill, however, the securities law says that is wrong.

Philosophically is it the right way? I don't know, but there is a legal basis for the prosecution.

Comment: Re:Algorithmic trading? (Score 2, Interesting) 299

Almost.

I can't speak for all of HFT, but for the most part they would be more than happy to get filled, but they don't get filled because they move too fast.

They are calculating a theoretical price and placing orders that would theoretically be good to fill and that are very close to or at the "inside" (best bid or best offer). The inputs into their theoretical model change frequently and they cancel and replace their orders frequently, most often before they have a chance to get filled.

The difference is that for the duration of the order (even if it is small) the HFT wants to get filled and isn't trying to necessarily push the price around. They know the likelyhood of a fill is small, but in a way they still "intend" to get filled and don't have the goal of manipulating the price.

The guys who got in trouble put in orders that they hoped would not fill in order to cause another automated participant to follow along, and then quickly cancel their own order and hit the target order.

Comment: Re:Algorithmic trading? (Score 4, Interesting) 299

>>So these guys figured out how to second-guess somebody's trading algorithm. How in hell is that a crime?

Not quite.

If they had figured out how to predict where somebody else's algorithm was trading, and trade against it for profit, they would not be in trouble.

What they did was figure out how somebody else's algorithm would react to stimulus, then entered created that stimulus, then traded against the result.

They entered orders that had no intention of getting a trade (and indeed would have been unhappy to have traded because they were unnaturally high bids or low offers). These orders gave the impression to both people and software that the market had changed for real. The algorithm followed the "fake" data and made too high of a bid or too low of an offer. They then cancelled their "fake" orders and instantly entered real orders on the opposite side to hit the algorithm.

This has been going in (sans computers) for decades, and is illegal in most regulated markets.

It is similar to the idea of leaking fake news and trading against the move and then making a profit when people figured out it was fake.

Consultants are mystical people who ask a company for a number and then give it back to them.

Working...