OUT OF CONTROL

[mawilson 2,133]

Ah, so someone is paying these billions of dollars to HFTs.

Sure, a few cents out of your $7 trade and other market participants who take liquidity

instead of providing it.

When you want your shares "right now" and you don't care about the cost of crossing

the spread (a penny for most liquid stocks), HFT trading algos are there to sell

you the shares you want.

[scandal 951]

Can you substantiate that? I still see online trading applications that average investors would likely use talking about $7.00 / trade. That could be more or less than the percentages you quote based on the value of the trade. Tell me this, though, when GS makes billions of dollars from their HFT operation, where does that money come from? Out of thin air? Or do less sophisticated market participants effectively pay for that?

I posted this a few days ago. It references some recent studies of market activity over the highly volatile weeks in early Aug and suggests that HFT may indeed have been a net-liquidity-adding factor, and net-calming factor on the high volatility days. In the sense that algos will often rush in to buy what panicked investors are selling, I can certainly see that being the case. A lot of algos are looking for reversion to mean in statistical arbitrage, and they'll supply liquidity to oppose a sharp market move in the opposite direction.

mawilson is quite right that spreads today are a minute fraction of what they were. Liquid US equities (think AAPL, QQQ, SPY etc.) are penny wide, and cash FX even tighter in today's micropip era. I can remember (and I'm old but not that old) prior to decimalization, and even before teenies (1/16), when typical spreads were 1/8 or even 1/4 (that's $0.125 or $0.25). Retail investors benefit from that decreased bid/offer spread. If you took out the HFT and institutional players (e.g. the GS that you like to demonize) you would have thinner markets and go back to wider spreads. Technology has made the game faster. That means only players with colocations and fast data lines and racks of high end servers even have a chance at playing HFT. It means I would never want to be a retail player stacked against them, day trading. But I don't mind as a retail investor having the liquid markets and tight spreads we get as a benefit of the huge trading volumes HFT pumps in. That's a net plus.

As to your questions of "where does the money come from? Out of thin air?" That's the age old question of the market place. It long predates HFT, and really has nothing to do with HFT. During bull markets all the longs are getting rich- big, small, institutional, retail. During bear markets the shorts get a turn. During quiet markets, anyone holding delta neutral strangles or butterflies gets clobbered, during volatile markets they're in the pink. So it goes. Markets "create" wealth and also take it away. HFT is just a variation on this long standing truism.

[Mr. Big Dog 10,747]

Sure, a few cents out of your $7 trade and other market participants who take liquidity instead of providing it.

When you want your shares "right now" and you don't care about the cost of crossing the spread (a penny for most liquid stocks), HFT trading algos are there to sell you the shares you want.

That still is an additional cost without any conceivable value. Whatever HFTs shave off, someone else pays. It's a zero sum game after all so someone's gain is necessarily somebody else's loss. HFTs do not actually produce any good - i.e. they do not add any real economic value. They make money from money, they take a cut from some other party's deal - that's all they do. They get privileged access to and hence get to manipulate prices. Not a dime of real economic value is created. The only thing that's created is profit. At someone else's expense.

[Mr. Big Dog 10,747]

I posted this a few days ago. It references some recent studies of market activity over the highly volatile weeks in early Aug and suggests that HFT may indeed have been a net-liquidity-adding factor, and net-calming factor on the high volatility days.

The key word is MAY. There are studies out there suggesting that CO2 emissions MAY not have any impact on the global climate. That said, there are also studies out there that found just the opposite - both for HFT and volatility correlation and CO2 emissions and climate change.

Fact is that there is more and more frequent rather than less volatility out there as the HFT strategies have taken over 70% or so of the equity trading volumes. Square that circle.

[웃 910]

About a month and a half ago I was trading SINA...Pretty exciting stock. Was $80 at the time and working with about 5 to 7 million volume a day.

I went long at $89.7 or something around there. Set my stop loss at $89.49 because it bounced and continued to bounce off of $89.50 which was the low for the day.

"Long" story short (lol) I'm sitting there watching it go to $89.50 waiting for the next millisecond to pass where it hits $89.49 and my market order fills between $89.47 and $89.49 (my guess for such a liquid stock).

You watch it stay at $89.50 for about 20 seconds and it slowly climbs back up to the .70s, .80s....I'm kinda sick of holding it and hoping it fills so I can take the loss and be done.

About the 3rd or 4th time it came back within 1 freaking penny of my stop it finally hits sub $89.50 for a new low.

I see a massive crash and see that my order filled at like $88.70s...It came out to several hundred down because I probably had 300 to 500 shares.

It was pure bullsh!t and just completely unlucky and a rare circumstance. I could see a $500 stock like Google dropping 80 cents on one additional share of volume but not an $80 stock with 1 to 3 penny spreads.

I brought this up cause you guys were talking about HFT and wish they didn't disappear and pull all orders off the table at exactly the break of $89.5 which was a decent place to set the stop.

Edited August 24, 2011 by Lord Infamous

[mawilson 2,133]

That still is an additional cost without any conceivable value.

Do you remember how people used to call their broker if they wanted to make

a trade and the broker would turn around and relay your order to the

trading floor, where a runner would relay it to the floor broker, who then

executed the trade; the runner relayed the trade confirmation back to

your broker who would tell you how it went and of course charged you a

hefty sum for his services?

Are those the good old days you are pining for?

Electronic trading changed all that. Now a retail investor such as LI or

yourself can pay $7 to buy or sell any reasonable number of shares in an

instant, from your web browser! You don't see any added value in that?

[mawilson 2,133]

I see a massive crash and see that my order filled at like $88.70s...

Why did it get filled at 88.70 when you had a stop at $89.49? Did you have a market stop instead of a limit stop?

Edited August 24, 2011 by mawilson

[scandal 951]

The key word is MAY. There are studies out there suggesting that CO2 emissions MAY not have any impact on the global climate. That said, there are also studies out there that found just the opposite - both for HFT and volatility correlation and CO2 emissions and climate change.

Fact is that there is more and more frequent rather than less volatility out there as the HFT strategies have taken over 70% or so of the equity trading volumes. Square that circle.

You asked

Can you substantiate that?

So I cited a relevant study. I agree with you that the verdict is out on the net effect of HFT. But I hear you claiming that HFT is responsible for all sorts of evils, which is also by no means a universally agreed position. Keep an open mind.

Is there more and more frequent volatility out there? I'm not sure. Maybe. VIX is hitting all time highs, that's true. But VIX didn't exist in 1987 or 1929 or other "high vol" eras. On a percentage basis the broader indexes I believe had their most dramatic swings in Oct 1987. More than after 9/11 or Sept 2008. In between the big moving days, we have plenty of slow churning days too, and that's in our current HFT era. That's how I square your particular circle.

Besides, arguing against HFT is a little bit like arguing against cell phones or automobiles or any form of modern technology. We live in a technological era, and humans will always try to use technology to do what they did yesterday a bit faster, a bit more efficiently. You risk being labelled a Luddite by railing against that darned newfangled technology. Better to adapt to it, see its benefits and curb its weaknesses, and find ways to leverage it, than to simply complain about it.

[웃 910]

Why did it get filled at 88.70 when you had a stop at $89.49? Did you have a market stop instead of a limit stop?

I had the stop that immediately does a market sell as soon as it touches the stop price ($89.49).

My petty 300-500 shares did not trigger the sell off obviously....It was not a high risk move and you wouldn't think this would ever happen. But it did. I mean 10 cents, sure, who cares. But this was just wrong!

I believe (not 100% sure) that the following day the stock shot up like $20 in one day...Completely out of the blue. Can't remember if it was on earnings or an upgrade or something else. Markets were flat and SINA gapped up about $5 and just kept going and going....So it was a double slap in the face.

ETA: The stop I am talking about is the only stop I have ever used and I like it...This case was rare..

Edited August 24, 2011 by Lord Infamous

[scandal 951]

It's a zero sum game after all so someone's gain is necessarily somebody else's loss.

But it's not a zero sum game. Who told you it was? On a long term basis the market will reflect growth in the economy. GDP goes up on average 2-3% annually over the long term in a mature economy like the US, and the stock market goes at double that clip due to productivity growth, market share growth, etc. In such a market all long term investors will make money. So it's not zero sum at all.

Even in the short term it need not be zero sum. Yes, for every buyer there's a seller. But investors have different expectations when trading. A hedger who buys an option to cover his position as insurance and has it expire out of the money doesn't mind his loss as it's the premium he pays on his insurance policy. Meanwhile the seller of the contract has earned that premium. It's a trade in the market place with a buyer and seller, a "winner" and a "loser" but effectively really only two satisfied counterparties.

[scandal 951]

I had the stop that immediately does a market sell as soon as it touches the stop price ($89.49).

My petty 300-500 shares did not trigger the sell off obviously....It was not a high risk move and you wouldn't think this would ever happen. But it did. I mean 10 cents, sure, who cares. But this was just wrong!

I believe (not 100% sure) that the following day the stock shot up like $20 in one day...Completely out of the blue. Can't remember if it was on earnings or an upgrade or something else. Markets were flat and SINA gapped up about $5 and just kept going and going....So it was a double slap in the face.

ETA: The stop I am talking about is the only stop I have ever used and I like it...This case was rare..

So you had a "stop loss" order, as you said, and not a "stop limit". A stop loss can indeed behave as you learned here - like a hot knife through butter. If there are no support levels below your stop it can go crashing through. When you trade, do you look at the Level II to see the size levels below the BBO? As many found out in the flash crash on May 6 201, or as people in thin stocks find out all the time, size can evaporate right below the top. Thin stocks in particular are subject to being manipulated in this manner. I'm not calling SINA a pump/dump stock - I know nothing about it. But it's not a big well traded name, that's for sure.

LI, you're trying to time single issue high momentum stocks intraday on technical indicators. I can't tell you risky I think this style of trading to be. Very, very risky. Very poor risk/reward profile. We've had this conversation before, I know.

What typical fast-trading HFT programs do, particularly "statarb" (Statistical arbitrage) models, is to trade actively in a large basket of stocks, say 200 or 400 or even 1,000 names. They'll often pair them up and do long/short combos, or they'll trade them in "clumps" with a net flat exposure (i.e. buy 300 ABC, 400 DEF, sell 200 UVW, sell 500 XYZ) in a weighted combination. The holding times are typically seconds or minutes, they're invariably flat end of day and they'll turn over entire positions 1000's of percent daily. They're looking for the minutest of edges, some belief that if ABC and XYZ negatively correlate then buying one and selling the other based on a momentary signal will profit. They expect results of the order 51-52% winning trades, 48-49% losing trades. That's enough with the volume of trading they do to justify the expenditure in trading infrastructure. They rely on super-fast market data feeds and execution engines, and on smart routing to guide orders to the best possible execution venues. With all of these models arrayed against you and the sheer volume they'll collectively trade in any direction, and the historical backtesting that goes into all of these models, you (or me) and our puny little stop-loss is like waving a red flag in front of a bull. You will get run over.

The only style of trading I think a retail investor has any business engaging in is long term bottom-up fundamental stock picking (if you have time and inclination) or long term buy-and-hold of well diversified efficient index funds/ETFs.

[웃 910]

I'm not calling SINA a pump/dump stock - I know nothing about it. But it's not a big well traded name, that's for sure.

SINA is huge. Absolutely enormous (ok, slightly exaggerating). It's on a level like Apple in terms of popularity (closer to BIDU...ok less but still very popular). And only getting more popular with time. Look at today. We'll call it a $97 stock with 7 million volume (trades going through each minute of the day, block shares exchanged, volume spikes, etc.). Type of stock where you could put 50,000 share block in and not budge the price more than a 25 cents....Total = $679 million dollars exchanged on the day.

CAT (well traded name, right?)....$83 and 12 million traded = $996 million

KO = $69 and 12 million traded = $828 million

X = $27 and 11.4 million traded = not even $308 million and nobody is gonna say that X is not a well traded name

F = $10 and 53 million = $530 million

BIDU = $136 and 10 million = 1.36 billion dollars exchanged today

Anything trading well into the hundred millions is very liquid. A couple hundred shares of SINA when it is trading a million shares an hour should have never gone from highest bidder $89.49 to $88.7..It happens but is extremely rare (usually the orders will get eaten up and it will drop fast penny by penny within seconds. But the stop order should fill on the way down and be relatively close to where it was set to sell.

Do any market buy or market sell on SINA at any time in the day and it will fill within 2 pennies of the ask or bid...This was just freaking odd!

LI, you're trying to time single issue high momentum stocks intraday on technical indicators. I can't tell you risky I think this style of trading to be. Very, very risky. Very poor risk/reward profile. We've had this conversation before, I know.

I agree with you. It is nuts. There are successful traders out there. I am not one. I "might" try again next January.

Edited August 24, 2011 by Lord Infamous

[웃 910]

5 Minute Chart of SINA 9:30am to 4:00pm.

In section with volume I have 2 horizontal red lines. The first one is 50k volume and the second is 100k volume.

The smallest bar happened in the 5 minutes from 12:25 to 12:30pm and traded almost 19k shares. That was by far the lightest trading of the day and most illiquid.

The heaviest was 247,000 shares in 5 minutes.

It consistently traded over 50,000 shares each 5 minutes through 12:25pm (and then again starting at 12:40pm)...

Opening 5 minutes = 212k volume

I'm just bored that is why I am making this reply

[scandal 951]

I'm just bored that is why I am making this reply

That's ok. We all do that sometimes.

[Mr. Big Dog 10,747]

Besides, arguing against HFT is a little bit like arguing against cell phones or automobiles or any form of modern technology.

Or maybe it's more like battling cancer.