Caroline Ellison, CEO of Alameda Research, pleads guilty to seven offences

Presumably i bought because I thought it was worth more than $100, it would only make sense for me to sell at $80 for a reason other than the price: fundamentals change, some bad event happening. If the only thing that changed is the price, I want to buy more at $80!

A loss of 10 percent necessitates an 11 percent gain to recover. Increase that loss to 25 percent and it takes a 33 percent gain to get back to break-even. A 50 percent loss requires a 100 percent gain to recover and an 80 percent loss necessitates 500 percent in gains to get back to where the investment value started.

Stop-losses are one mechanism for implementing that, but they are crude, and big players don't actually use them. In that regard, she was correct. For one thing, if you trade size, you don't want to dump everything with a marketable order at once, you have to scale out even if you are wrong. Secondly, some strategies will always perform worse if you add stop loss levels, notably mean reversion. But they can make sense, for instance, for trend following, because there are clear levels at which you can say that a trend has ended. Lastly, high-frequency traders like market-making firms, which Alameda was supposed to be, don't use stop loss orders, because they are not trying to take directional bets, they are just trying to capture the spread while keeping their book balanced.

Also, it's a psychological thing: traders think in risk. If they were mistaken with a trade idea, they close the position with a small loss. Easy, just move on to the next!

But when a position is already deep in the reds, it becomes increasingly harder to press the "close"-button and realize the loss.

I know somebody who lost a few hundred grand just a few weeks ago because he was unable to close the position. He didn't set an initial stop loss, and then started believing in his trade (or the company) too much, although the market was telling him otherwise. He started doing dollar cost averaging and other stupid stuff. Actually, this guy is incredibly smart, but became too attached. I feel genuinely sorry for him. But that's what happens, when you don't have the discipline.

A stop-loss-limit is better in that it makes the potential consequence of a stop-loss more clear: You could set a stop loss limit order with a limit of $0 to create a standard stop loss-- making it explicit that you're willing to potentially sell the asset for $0/share: which is what a stop loss is willing to do. (+/- market circuit breakers, which generally don't exist in cryptocurrency markets.)

If it's not immediately apparent why stop losses are a hazard: When markets are volitile the supply of standing orders near the spread tends to thin out-- for some assets, like the worthless magic beans FTX and friends specialized in owning, the markets are never particularly thick. What a stop loss order will do is once the market ticks below your threshold it will dump into a market order. Market prices are not continuous. If someone sold at $100 that doesn't mean you can sell at $100-- a market sale might be at $80-- locking in a substantial loss that otherwise would have been a momentary blip and never impacted you. These orders essentially automate one of the worst practices of inexperienced investors that result in loss: panicking at every dip and selling at a loss when nothing fundamental has changed.

What joe-sixpack thinks a stop loss will do is guarantee him a floor price. It will not. Joe could buy put contracts to create a guaranteed floor price but they cost money-- that cost is a direct sign of how much stop loss orders do not work.

A stop-loss-limit at $100/$100 would do what was expected if it executes but it usually won't execute except when the price dips and then recovers-- the case where you would have preferred to have your stop loss not exist at all. Seldom do people want a "Sell my stuff at $100/share if the price gets under $100 but only when it recovers"-- that would probably only be justifiable to the extent that the drop showed your thesis about the investment was wrong. Fundamentally the guarantee people want here can only be had at a price, and paying that price is a reasonable part of risk management.

If you're trading very small amounts of very liquid items on highly surveilled and regulated markets then perhaps you can get away with using them without getting too greatly burned. But at the same time puts for the same assets are usually fairly inexpensive. In the cryptocurrency sphere it's just not that unlikely that 'exchanges' (particularly bucket shops like FTX where the exchange is substantially the counterparty in the activity) has some script that counts up all the users stop loss orders and figures out how much profit they could make causing a momentary blip in the trading price just to trigger them. There are plenty of people in the industry that don't believe it would be unlawful to do so, seeing as how the traded assets aren't securities.

(and IIRC long before FTX's collapse there was a lawsuit alleging that they engaged in that kind of manipulation)

So mocking the stop loss comment seems a bit misplaced, but it's worth noting that the question was really about risk management and she didn't give a useful answer to the intended question either-- especially since the rubbish they owned was hard to impossible to risk manage and for good reason.

A stop loss could be set at a threshold that isn’t so tight to lose money on normal variations, low enough if it’s acceptable to take the loss and and capture and deploy the capital elsewhere, and not too low to never be triggered.

She's a freak in a freak show. The praise is ironic.

https://caroline.milkyeggs.com/

Kind of assumes markets yield truth when left alone. It's equally true that a small but well-coordinated swarm can shape market behavior when directed at the right target, even if the swarm's activity is not correlated with anything like underlying asset value or earnings.

I am not a fan of WallStreetBets but I don't think of them as idiots; more that they're making the argument (by demonstration) that their open and shameless manipulations are just a less polite version of what already happens behind the respectable facade of financial institutions staffed by well-credentialed traders in nice suits. Having spent a lot of time around trading desks (from tiny funds to private banks) I don't think they're wholly wrong. Although they epitomize the greatest weaknesses of crowd/swarm behavior as well its strengths, they are typically playing with their own money rather than collecting 2% to do the same thing with client funds.

We don't share the same concept of humor then.

Super funny to every single person who lost money to the person "getting in the record books for 'most money lost by a single individual'".

I laughed.

Putin would beat that easily if you add lossess to business and business opportunities lost

I encourage you to watch Coffeezilla's video about Tether. There's lots of shady stuff going on behind the scenes, like them borrowing $383M from Bitfinex and showing that to an auditor to prove they were solvent (~20:00).

https://youtu.be/-whuXHSL1Pg

A lesson I've taken from the last few years is that consequences can take a very, very long time, but that doesn't mean their strategy is working even if the tactics seem to be succeeding in the short term.

It doesn't stop there: that lawyer at Bitfinex you mentioned was working back then at Excapsa with Dan Friedberg, the top lawyer at FTX.

IE, if 50% of the principal was spent on crypto gambling and Tether lost ~80% of it (so they lost 40% of the total principal), and the other 50% was put in safe assets and they got 20% of that (10% of the principal), they'd be down to 70% of the principal. But they're only going to be insolvent if people try to cash out 70% of the Tether in circulation at once. And like you said, just putting that into fairly safe investments is going to undo a lot of the damage. Heck, even just getting the returns an online savings account is giving now would have them make back the principal in about a decade.

Which isn't to say they won't fall apart. But I think a lot of people go to far assuming that their collapse is inevitable.

Conspiracy requires at least 2 people.

So for example if two mates talk about that they are going to kill person Y, that is the agreement. And then one of them buys a gun, that is an overt act. And even if person Y has very good security, and the two dudes have no chance in hell of murdering Y ever, that is still a “conspiracy to muder” charge.

It's a way to let a prosecutor get ahead of an unsuccesful, and basically a freebie in terms of sentence amplification.

Now, a wiretap of a boss ordering a hit could result in the same sentence as the gunman.

Unfortunately it has been abused greatly, including drivers who never got out of the car sentenced to death for a 7/11 robbery gone bad, etc.

Further, as you can imagine, it gives the prosecution massive (many would say unfair) leverage over associates to a crime, given that they could face the same, full sentence if they don't cooperate. It's a major reason behind the enormous percentage of criminal charges not going to trial.

So they did have the "smarts", it just depends on if this started out as, or became, a massive fraud.

This is not money, not on a bank with FDIC, and not on a brokerage with SIPC. If you don't think so, I've got some usdb on realmoneyexchange.com to sell you.