• C++ Programming for Financial Engineering
    Highly recommended by thousands of MFE students. Covers essential C++ topics with applications to financial engineering. Learn more Join!
    Python for Finance with Intro to Data Science
    Gain practical understanding of Python to read, understand, and write professional Python code for your first day on the job. Learn more Join!
    An Intuition-Based Options Primer for FE
    Ideal for entry level positions interviews and graduate studies, specializing in options trading arbitrage and options valuation models. Learn more Join!

Clamping Down on Rapid Trades in Stock Market

Joined
1/10/11
Messages
124
Points
38
Regulators in the United States and overseas are cracking down on computerized high-speed trading that crowds today’s stock exchanges, worried that as it spreads around the globe it is making market swings worse.
The cost of these high-frequency traders, critics say, is the confidence of ordinary investors in the markets, and ultimately their belief in the fairness of the financial system.
“There is something unholy about them,” said Guy P. Wyser-Pratte, a prominent longtime Wall Street trader and investor. “That is what caused this tremendous volatility. They make a fortune whereas the public gets so whipsawed by this trading.”

http://www.nytimes.com/2011/10/09/business/clamping-down-on-rapid-trades-in-stock-market.html?_r=1
 
In Europe they're talking about levying a small transaction tax on trades -- I suspect this will probably kill off algo trading (but I could be mistaken).
 
I'd suggest mandating a minimum limit order persistence time and stopping sub-penny orders.

None of that will prevent crashes, though; it will just level the playing field a bit and cut down on some of the more egregious games that are played. The fact is that any trader, human or machine, will pull bids if panic strikes. If nobody has the responsibility to make a market, then there is always some threshold of fear past which no market will exist.
 
I think it'll just increase the spreads. So long as there is liquidity a predominant chunk of it will be electronic.
 
Yes, but tighter spreads don't mean lower transaction costs. A case in point:
http://www.nanex.net/StrangeDays/08022011/ELNK_08022011_1.jpg

ELNK_08022011_1.jpg


Woe betide the poor shlub who has to click the "Place Order" button to hit a bid here. It fluctuates by 2.5% percent of spot about every 2 seconds.

I have nothing against electronic trading per se, just rules that allow it to fleece people under the guise of "providing liquidity."
 
Bob, this kind of abusive behavior should be taken care of by SEC/Finra with one phone call. They have the data, and they can easily find out who was the idiot-quant trader that thinks this is a way to "provide liquidity" . If regulators are not competent enough to stop it, they should subscribe to NxCore and hire a few youngsters to look at the "pretty" pictures.

Dark pools are the ones to be closely watched. Trade-at rule would be a step in the right direction; hence no need to mess around with sub-penny trades.

The reason you have tight spreads is because there's a guy willing to stick his neck out and show you a bid. There is no reason why HF trader can step in front of the displayed liquidity when a person is right and let you be hit when one is wrong. Liquidity providers should be rewarded for putting their head on the chopping block.
 
we can always go back to having 200 drunk guys on an exchange floor shoving paper tickets in each others faces. At least it's entertaining to watch.

A minimum limit order persistence time will solve nothing for the people that complain about HF now. They will still be too slow to react compared to the HF systems they claim screw them over day in and day out.
 
Impose a limit on the ratio of orders submitted to actual trades, then fine people who go over their limit (measured on a daily basis). Would cut down abuses and reduce bandwidth issues.
 
we can always go back to having 200 drunk guys on an exchange floor shoving paper tickets in each others faces. At least it's entertaining to watch.
The fact that the system in the past was neither fair nor efficient does not constitute a defense of a later, different system that is also unfair and inefficient.

A minimum limit order persistence time will solve nothing for the people that complain about HF now. They will still be too slow to react compared to the HF systems they claim screw them over day in and day out.
I suppose it depends what the limit is, doesn't it?
 
... but not at all entertaining!
It just needs a good UI--something that involves a joystick and lots of movement / explosions. Do it right, and pre-adolescents will be making (losing) millions on HFT strategies from their XBoxes.
 
Impose a limit on the ratio of orders submitted to actual trades, then fine people who go over their limit (measured on a daily basis). Would cut down abuses and reduce bandwidth issues.
This is already in place, at least on CME.
 
The fact that the system in the past was neither fair nor efficient does not constitute a defense of a later, different system that is also unfair and inefficient.
Bob, while I generally respect your opinion, you are going to have to provide a concrete argument for why the system is unfair and inefficient. Otherwise, it just sounds like conspiracy theorizing.
 
Bob, while I generally respect your opinion, you are going to have to provide a concrete argument for why the system is unfair and inefficient. Otherwise, it just sounds like conspiracy theorizing.
I'll let two guys with ten years experience at Instinet and a brokerage firm of their own make the argument better than I could:
A good interview that gives some background:
http://www.themistrading.com/article_files/0000/0576/61110_Welling_Weeden_PlayingFair.pdf

A recent letter tracking progress (if any) since the flash crash and recommending changes:
http://www.themistrading.com/stories/0000/0027/042511_Analysis.pdf

An excerpt from the latter:
What have we learned?
1. The nation discovered that our markets have changed drastically. While market players and insiders find this fact to be obvious, it was a wake-up call to the investing public – the hundreds of millions of regular folks, who save, invest and own capital, as opposed to renting it.

2. Investor confidence was substantially eroded. Trading volume plummeted and money left equity mutual funds en masse week after week.

3. The markets cater to high speed gambling and traders, not capital formation and economic growth.

4. Volume does not equal liquidity.

5. Our own government agencies and regulatory bodies have had a large hand in helping to create this fragmented “Franken-Market” structure.

6. Our own government agencies and regulatory bodies have been slow to change and correct unintended consequences because of:
• Budget issues and a lack of industry experience at best.
• Cozy relationships with the firms and insiders that have pushed for the current structure at worst.
 
Interesting article.

While I agree about the erosion of confidence due to events like the flash crash, a lot of what they say after that sounds like conspiracy theory with some truth laced in.
 
For once I see a reasonably cogent argument. Seems like they are arguing more about conflicts of interest, investor confidence, market fragmentation etc than whether or not people should be allowed to trade quickly. I don't disagree with the general thrust of any of this, but again that doesn't amount to a condemnation of trading fast.

So I don't see how the minimum limit order persistence time is an optimal solution unless it was applied to DMMs only. Fact is, nobody wants to stick their neck out, except DMMs who get privileges for doing so. This overall idea of "nobody should trade too fast" is too much of a kludge.
 
I agree that speed per se is not the problem, as long as what's actually taking place is trading. In cases like the "pretty picture" pasted in above, though, those oscillations in the NBBO are not arising from trading. As near as I can tell, it's an effort to create adverse selection in the execution of market orders. Those who justify adverse selection on market orders on the basis of such orders draining liquidity should then agree that charging a fee to cancel limit orders is also justifiable on the same basis.

As for the question of fairness, if we stand back from it and evaluate it by common, nontechnical standards, I think in basically any transaction a significant difference between the price you are apparently being offered and the price you actually pay qualifies as sleazy at least, and unfair if there's no ability to opt out and no physical way you could have said "yes" faster.

I understand a lot of finance runs by siphoning off retail money in many different, inventive ways, but in a world where equity long-term returns are zero and bond yields are zero, too, those in the fund, brokerage, and exchange businesses might give some thought to what happens as increasing numbers realize they can make more money (or, rather, lose less) by stashing cash under a mattress than they can by entrusting it to the "professionals." This realization may be particularly damaging in an environment where ever-growing numbers of retirees have to convert investment assets to cash to pay the bills.
 
As for the question of fairness, if we stand back from it and evaluate it by common, nontechnical standards, I think in basically any transaction a significant difference between the price you are apparently being offered and the price you actually pay qualifies as sleazy at least, and unfair if there's no ability to opt out and no physical way you could have said "yes" faster.
This is exactly what Marketable Limit Order IOC addresses, and this specific order is indispensable for HF arbitrage.
 
The question I have to ask is is that picutre the result of one algo or two? Because as we've seen the strategies in algos vary and respond to the limit orders placed by other algos. Another pretty (and related...) picture would be that book that cost 34 million on amazon... I don't think that that was intentional or malicious. But hillarious - sure.

As far as retirees go - the average savings of them is $1 mil+. The median savings is $80k. They don't really have much to liquidate...
From the standpoint of someone who wants to transact in the market, does it really matter whether it was one, two, or eighty seven algos that produced that cycling behavior?

As for the second point, here's a link to paper by the SF Fed giving a demographic analysis that projects stock prices will stay below the 2010 levels for the next 15-20 years:
http://www.frbsf.org/publications/economics/letter/2011/el2011-26.html
 
As a side note, according to FT article from today: a study by CFTC claimed that the traditional market makers bore more blame for the flash crash than HFTs because they never stepped back in, whereas HFTs came back after liquidating their net positions.
 
What sort of idiot would jump back into danger when they had no idea what the cause of that danger was? To blame traditional market makers because they DIDN'T all pile in to solve someone else's screw up seems hilariously cheeky.
 
Back
Top