The death of short term retail trading?

Trading Discussion
1

I’ve had a lot of time away from being involved in the markets actively, since I’m going through business school.

In the midst of my studies, and reading trading forums, I’ve done some rumination over the subject: is it ever possible that retail traders will not have the opportunity to exploit market inefficiencies anymore, especially in a short term time frame?

One thing we’re seeing a lot of is algorithmic trading, especially high frequency quantitative trading which heavily cuts into day trader territory. The trend is apparently becoming seriously noticed enough to the point where trading desks demand quantitatively oriented hires, and big prop firms are coming to blows against the algo boxes.

One such example is Schonfeld securities, a highly profitable prop firm which announced a layoff of a slew of day traders.

As it said in Schonfeld’s email to employees as it was announcing layoffs: [I]“It is getting much tougher for traders to make a living or get by. The direct competition from [U]black boxes, stat arb and high frequency trading[/U] which continues to grow at exponential rates is here to stay and has caused us to change our outlook for lesser skilled traders.”[/I]

Another interesting point was made in a Bloomberg article on the same subject: [I]"“There is no doubt that manual point-and-click-type day traders are at a disadvantage,” said Paul Zubulake, a senior analyst at Aite, a Boston-based research firm. “It just is that if you are in a marketplace, especially arbitrage, if you are not trading at a high speed, you cannot compete, you have no chance. [U]A few years ago you could maybe survive by seeing something visual, but now the machine will take out the price inconsistency very quickly.[/U]”"
[/I]

Another such comment comes from Mike Bellafiore, an owner of an equities prop trading desk (SMB Capital) who says in his blog “…in this day and age of HFT battling HFT battling third party HFT for a half a cent, the Scalp Trade has been deemphasized on the best intraday trading desks. The programs are just too fast programming out many of the easy traditional scalp plays.”

While equities so far have the highest HFT volume, Forex is second in line along with futures. A friend of mine who is an FX dealer told me that there’s an increasing move towards replacing dealers with algo boxes.

I downloaded a reputable book on quant trading, figuring that maybe this is something that I should begin developing my skills in, but the econometrics they throw at you there are only something a math major would be able to hack. Another issue is that many of the companies - on top of having access to top quant thinkers - use black boxes use very high powered computers that have extremely rapid connections to the exchanges, so it makes it hard to truly complete.

On the other hand, you have thinkers who believe that while markets change, they also stay the same. Jim Rogers said this in Market Wizards: “[B][I]Do the markets behave any differently now because so much money is being managed by trend-oriented systems?[/I][/B] [I]No. They may not always have been on a computer, but there always have been systems. I guarantee that you can go back 100 years in the market and not find a single decade where there hasn’t been some kind of system, some kind of new formula developed to play the markets.[/I] [I][B]So the markets today are basically the same as the markets in the 1970s, 1960s, and 1950s.[/B]
The same as the markets in the nineteenth century. The same things make markets go up and down. They have not changed the rules of supply and demand.”[/I]

Granted, this was before the HFT craze…

Now if there’s one thing I’ve learned in business, its that the competitive landscape is always changing - traditional business models face innovation and then you either change or die. Sometimes, certain categories of competitors gain the upper hand in markets, i.e. there are those who manage to erect a high barrier of entry which can prevent others from competing. A perfect example of this is a Wall St firm which develops an HFT black box, they have capital and resources that small traders (you and me) do not have, and unlike before, they can also trade with small size - HFT is all about taking pennies out of the market at a time.

So, the point is: is there ever a risk that a class of competitor will emerge that will disable short term intra day traders from maintaining an edge in the market? Has that risk arrived yet or no?

2

Well, only if you are scarping i guess…

A trend is still a trend, so trend trading should be fine,

3

Hello TonyIommich,

Well I have to say that I’ve read your post more than once and it is a post that at very least deserves recognition!!! Excellent!!! Thank you.

I need to think about the content of your post at length before making any comments. One thing that does immediately come to mind though: are these recent (and ever increasing number of) ‘(mini) flash crashes’ not a result of HFT algorithms??? Because if so: I’d not be TOO sure newborn1000 that trend trading should be fine i.e. those ‘flash crashes’ are 100% guaranteed to take out your stops if you’re on the wrong side of the ‘flash crash’.

Regards,

Dale.

4

The response I would have to this is that markets need to be looked at seperately where HFT is concerned. It’s one thing to talk about the stock market where the impact of HFT (or it’s withdrawal) can be substantial - as we’ve seen - but a different thing to talk about something like forex where you’re talking about a much more massive market, but also one that is more disperate in terms of order execution. That’s on the micro level. On the macro level, there’s the question of how the algos and other cross-market strategies are creating increasingly high rates of correlation between markets. Definitely interesting stuff to ponder.

5

I was thinking: was this type of programmed trading not largely responsible for the October '87 crash (or are we talking about two different things here)??? Of course: it’s stocks that I’m ‘on about’ i.e. if I understand John correctly this HFT is more of an issue with stocks than forex is it not??? If I remember correctly: ‘circuit breakers’ were put into place afterward (which as I understand it have been largely done away with now UNTIL THE NEXT TIME)???

On a personal note: one would HOPE that if it became an issue then the SEC or the CFTC or some or the other regulatory agency would ‘step in’ to protect investors (was there not some issue or the other raised just a few months ago about this i.e. HFT giving an unfair advantage to ‘those’ firms over the ‘average investor’ or something like that)???

Regards,

Dale.

Edit: sorry but the issues you raise are fascinating so I’m going to keep commenting as things ‘come to mind’ if you don’t mind.

One other thing that I just thought of (while reading your post again):

Jim Rogers is quite right and I agree with him (who am I ANYWAY to DISAGREE with what ‘the man’ says or thinks) i.e. the markets are still the same i.e. ‘supply and demand’. HOWEVER: Jim Rogers does not ‘trade’ like we do??? If anything: in my opinion he is ‘tradevesting’ i.e. he looks at the long term fundamentals of a particular commodity and buys or sells based on those fundamentals. I doubt very much that he ‘trades’ like we do e.g. using stop losses and trying to ‘catch’ every turn of the market so for him (as I understand it anyway) it does not matter if there are a KAJILLION price fluctuations in a given day and how fast they happen i.e. he’s ‘in it for the long haul’. So yes: given HIS ‘style’ the markets are the same.

6

My feeling is that while the government might force certain HFT’s to classify themselves as market makers and always be in the market as liquidity providers, that will only temporarily stave the HFT craze. HFT it seems is here to stay, its not possible to ‘put the genie back into the bottle’ as one Bloomberg article noted.

However, the issue is that the market in the end moves because of fundamentals. Intraday inefficiencies may radically change, but the bigger picture won’t. Perhaps this means that swing trading might end up being a better bet for retail traders?

7

So called “program trading” was blamed as a culprit in the Crash, but as far as I’ve ever heard, it was more what the computers were doing rather than that it was computers doing it. Specifically I’m referring to “portfolio insurance” where futures and options were sold as the market fell to protect the downside of portfolios, which naturally pushed the market down further, causing more selling. That selling of protection, ironically, exacerbated the whole thing. We saw elements of a similar sort of cascade during the early parts of the financial crisis.

Of course: it’s stocks that I’m ‘on about’ i.e. if I understand John correctly this HFT is more of an issue with stocks than forex is it not???

That’s my current view on things.

8

Yes, this high correlation is making it very difficult to diversify trading. I check the intraday correlations between the S&P and EUR/USD, and its just amazing how closely they move. For a long time it was USD and JPY as risk averse, and EUR, AUD, NZD, GBP as risk - they all moved so closely together.

9

From today’s Wall Street Journal on commodity correlations:

"How on earth did sugar and stock prices get stuck together? Sugar, says Mr. Simons, is now both an “energy commodity” and a “growth story,” since much of the Brazilian crop is used to produce ethanol. That gasoline additive is linked to crude-oil prices, which in turn are sensitive to monetary policy and global economic growth—the same factors driving stock prices.

But there is another, less visible force at work, Mr. Simons says. [U]Algorithmic trading programs, or “algos,” automatically buy and sell a wide variety of assets based on mathematical models. An algo doesn’t know or care why two assets are moving together; it merely is programmed to recognize that they are doing so. As soon as a computer places bets that such a linkage in prices will persist, other traders—computers and humans alike—tend to take note and follow suit. That can be true, Mr. Simons says, whether or not a correlation is driven by fundamental economic factors.

“We’ve gotten to the Frankenstein point where algos are self-programming, and they evolve to chase these relationships,” Mr. Simons says. “That’s created a sheer wall of money that is forcing other people’s behavior into the same pattern.”[/U]"

10

Unless price movements become completely random, like a lottery draw, then I don’t see how retail trading can die. Short-term or long-term doesn’t matter. Long term moves related to fundamentals or public sentiment won’t happen without short term moves preceding them.

11

Its not the randomness that I’m concerned about, its the fact that markets may become increasingly more efficient. In other words, you won’t have the whole issue of overbought/oversold as much. Another problem, if the trading game shifts entirely into quant mode, buy/sell decisions will be harder to quantify because they are not happening through clear visual/numerical cues, but are based on econometric formulae. In other words, the price will be behaving in a new ‘code’ that discretionary traders won’t be able to decrypt as easily.

12

All big moves start small. Like a car, what is at first moving a millimeter until it is moving over miles. Plus, like a car, price stops, starts again, moves around and changes speed and direction on the journey to the destination. That, I assure you, will never change, whatever supercomputers or algos trade! Why? Because the market is a connection of two opponents. While one party tries to move the price up, the other party tries to move it down. If those parties are built by computers or humans or a mixture is just a question of style. Like chess algos. Do chess algos change the rule of chess? Or bookkeeping. Did computer algos change the rules of bookkeeping?

Plus let me tell you this: The market is not inefficient, it is in fact very efficient. Because it connects those two parties (bulls and bears) and their single entities in a perfect fashion. What makes it inefficient are human artificial rules.

There is the simple fact that values change and those values change because of human emotion. Price at markets are just a indicator for those values. Even if nobody would trade, those values would change the same way, because those values come out of humans emotions, not a price chart or a market.

13

Chess algos don’t change the rules of chess. HFTs actually do change the stock market trading rules somewhat for all participants. HFTs flood the stock market with massive quote-stuffing (sometimes 20,000+ orders a second) and can cause chaos to a particular stock price in just a few seconds. For example Progress Energy (PGN) was flash crashed from $44 down to $4 in literally a second in late September. This wasn’t a fat finger - it was an algo taking out all the bids and then cleaning up as it moved price back up to $44 again. All these orders that they submit they have no intention of ever taking - it’s basically price manipulation done at extremely high speed.

So if you’re a trader that happens to be in one of those stock prices that gets burned and your stops are hit then tough luck. There was absolutely no reason for the company to lose all that value in a second but this is the new trading game. This is why many day traders are on the way out of stock market trading - it’s become increasingly difficult to compete against HFTs and many firms are letting their prop traders go. In addition to all the quote-stuffing price manipulation that HFTs carry on they also handsomely pay the Exchanges to have their servers co-located to the Exchange’s servers so not only are they faster with their order execution they receive their prices ahead of everybody else too. Day traders are being forced to move away from the shorter timeframes completely in order to even survive and this means they have to change their whole fundamental understanding of the market in order that they can pick a direction and sit through whatever the algos do with churning prices intra-day.

I don’t know what effect HFTs will have on forex once they move in. The market is massive so it’ll be hard for them to impact particularly liquid pairs like they can do on the stock market. I think they’d still have an impact on very short timeframe forex traders - you won’t want to be going up against rampaging algos who are getting prices ahead of you and can execute 20,000 orders at the same time you can get one in.

14

I would actually argue two things.

First, in my view the more algos that attempt to trade the same patterns, the more likely they are to create imbalances. It happened with LTCM. It happened during the financial crisis. Granted, those weren’t day trading developments, but the idea is the same. There’s only so much liquidity.

Second, so long as the algos are following rules they offer the opportunity to be gamed - either consciously or unconsciously. Trading is all about pattern recognition. You watch the market enough you will learn to recognize the patterns that trigger the algos to move and the patterns created by their actions.

15

Hello,

As you know I think this is a great thread.

I thought I’d contribute this (just received from Larry Levin). Note the [B]boldface[/B] type:

[I]"Sell-Off[/I]

[I]I was asked by several family members and friends this afternoon what had happened to the markets today; why did it sell-off so much? Much of the following is well known, but I have tried to add a few extra details that have recently developed.[/I]

[I]What is scaring the market?[/I]

[I]North Korea attacked South Korea today and South Korea responded with something like, “If this happens just one more time, we will respond with full force.” Clearly, another war in the world will not be good for the markets.[/I]

[I]Ireland’s government is falling apart. Here’s the problem: citizens losing complete faith in government leads to hell in the streets…and this feeling is spreading in Europe.[/I]

[I]Portugal’s CDS spreads widened to its highest levels ever. This leads to much higher borrowing costs for Portugal, which it cannot do since it is broke, thus pushing it closer to bankruptcy.[/I]

[I]Spain’s CDS spreads widened to its highest levels ever. This leads to much higher borrowing costs for Spain, which it cannot do since it is also broke, thus pushing it closer to bankruptcy.[/I]

[I]A bankrupt Spain could bankrupt Italy, which will bankrupt the entire European Union (EU). The Euro currency could fail to survive.[/I]

[I]The insider selling case is even bigger than we originally thought. New mega-companies on Fraud Street are receiving subpoenas like; SAC Capital, Citadel, Wellington Management, and Goldman Sachs all named today.[/I]

[B][I]Next to be investigated may be the HFT firms.[/I][/B]

[I]Bank runs have been reported in Europe.[/I]

[I]Trade well and follow the trend, not the so-called “experts.”[/I]

[I]Behold the age of infinite moral hazard! On April 2nd, 2009 CONgress forced FASB to suspend rule 157 in favor of deceitful accounting for the TBTF banksters.[/I]

[I]Larry Levin [/I]
[I]Trading Advantage"[/I]

Regards,

Dale.

Edit:

I don’t know if THIS is relevant to the thread but ‘take a look see’:

Goldman to End Proprietary Trading - Yahoo! Finance

16

Awesome thread TonyIommich!!

I too believe this does and will have far-reaching consequences for us as retail traders - the “death of retail FX”, I don’t believe it’ll be that drastic.

I just posted an addendum to this thread where BIS has released a report outlining what is behind the increase in FX volume over the past 3 years.

Guess what’s at the top of the list??
Yep, increased activity by HFTs, summarised below.

  1. Greater Activity by High Frequency Traders
  • EBS Spot Ai was a key turning point for the algorithmic trading market
  • Introduced in 2005, it gave major institutions access to the inter-dealer market
  • Between 2007 and 2010, its share of trading on EBS grew from 28 to 45%
  • CME also launched an algo platform in 2002 and over the past 3 years, the turnover in FX volume has more than doubled
  • High Freq trading is estimated to account for 25% of spot FX activity

I’d recommend anyone interested to take a read of this very insightful report by the BIS for more info. Relevant pages are 27-40 here.

17

Although we do ignore it most of the times or take it on a very light way… like I read somewhere in the thread ‘I would agree if it is random, but…’

If you really want to know if it is or it is not, then you have to work it out honetly and methodically. And dont be surprised by the results. People just keep on trying various things and sometimes somethings work and sometimes some other things dont work, things which work (randomnly) come into notice, which dont (randomnly) go out of notice… its a complicated cycle but can be really tested with data if someone is willing to put his energies into it… for the rest they can just keep arguing, commenting, trying to sound intelligent, its a good game and timepass.

Mark

18

Going with the trend is to have less risk in trading.

19

Unless the algos/HFT get regulated!
You’re forgetting a simple point which is that even if all human traders were removed from the market, and all you had left was algos, there would still be a losing algo. Not all algos are profitable and they still suffer from similar problems that could be exploited.