Tuesday, September 25, 2012

Trade Ideas' Comment Letter to the SEC: Solving Cuban's Complaint

The Managing Partners and Board Advisors of Trade Ideas LLC submitted the following comment letter to the U.S. Securities and Exchange Commission after recent meetings there.  The letter is in response to the upcoming roundtable the department has organized to receive input on how the financial markets might be strengthened in light of the confidence-shattering events of the last 2 years.

Read the letter below and the references to the fine comments of others we make.  After ours I'd urge you to read remarks made by Mark Cuban, stock trader and owner of the Dallas Mavericks. They are particularly insightful.
September 10, 2012  
Securities and Exchange Commission
100 F St. NW
Washington, DC 20549-9303
Dear SEC:
Here are comments on market technology and stability issues that reflect the views of the Managing Partners and Board Advisors of Trade Ideas LLC.
Respectfully submitted,
David Aferiat    
The Value of Real-Time Monitoring & Assessment to Promote Self-Correcting Market Behavior and Efficiency 
Managing Partners of Trade Ideas LLC 
It has been described to us that the SEC acts as the police officer or detective who shows up at the scene after an accident to sort through the details, write a report, and, as necessary, suggest improvements to signage and driving behavior. For decades this approach has served the SEC well in its ability to establish rules of the market by which all participants would abide.
The U.S. equity market today, however, has become a new frontier to which increasingly innovative technology is unleashed. These new technologies threaten to add systemic risks, costs and increased instability to our market infrastructure. Up until the last 24 months the network linking various market participants together from around the world (e.g., bankers, investors, traders, brokers, agencies, media, regulators, exchanges, and others) handled millions of transactions with little to no impact and annually squeezed out costs and improved market efficiency.
In contrast today’s new technologies, like adaptive execution algorithms, practice ‘quote stuffing’: exponentially increasing the number of quotes, often well outside typical spreads, to generate a fog of war through which other competitive algorithms must pass and process as they seek to fulfill their execution objectives.  These technologies operate at increments smaller than nano-seconds and exceed human cognitive abilities to manage them without the aid of a new approach to monitoring market behavior. 
The SEC should tap key, real-time technologies to identify the traffic patterns most likely to produce accidents and respond to the scene more rapidly than previously to register what happened and who is responsible. Here are some suggestions of how real-time monitoring and assessment capabilities will improve market efficiency and the public perception of SEC as an involved, engaged regulator.
1.  Commit to real-time monitoring and assessment of market activity. 
Market crashes occur almost daily in today’s market. They only now reach the level of attention when multiple crashes occur at once and/or for a sustained period of time.  
The public perception of the SEC is that it regulates the excessive and illegal behaviors of market participants. Regardless of what the SEC’s interpretation of its mandate may be, its perception as a competent regulator is improved to the degree to which the SEC appears to be involved and capable of real-time monitoring and assessing what has happened and who is responsible as quickly as events occur
2.  Deal with message overloads with intelligent load shedding.
Our suggestion builds on top of the excellent advice proffered by James J. Angel ‘s letter to the SEC dated September 5, 2012. We agree with all of Angel’s suggestions. Below is his advice with respect to shedding load: 
We have recently seen many instances of “quote pollution” – huge spikes in quote volumes that do nothing to enhance price discovery. These excessive cancellations raise bandwidth costs for everyone, and can cause market-wide network degradation. 
Implication: We should have reasonable market-wide programs to reduce quote pollution, and mechanisms for load shedding in times of peak demand   
Load shedding should be prioritized by the source of quote pollution responsible for the most activity.  API channels, through which the vast majority of quote stuffing occurs, should be limited before retail channels like online and phone orders and other non-algorithmic order flow
3.  Harness real-time assessment and monitoring of structured (i.e., technical,
fundamental data) and non-structured data (i.e., news, social media) to watch for incipient market failures. 

Mr. Angel provides the following advice: 
Our market stability measures (circuit breakers) should be based on more than just price. In particular, we need three dimensional stability measures based on 1) price, 2) volume and liquidity, 3) data integrity. Disruptions in any dimension should trigger protective measures such as trading pauses or load shedding. 
We concur. Real-time assessment and monitoring allows for an understanding of these market stability measures and what their baseline behavior looks like.  A real-time statistical analysis determines what unusual looks like as events happen. Unusual activity across pairs such as price and volatility, volume and liquidity, and data integrity as measured by print and quote rates become alerts when one pair registers an extreme reading or a combination breach a number of standard deviations from baseline behavior. 
4.  Use sunlight as a disinfectant. 
Market participants will do anything to keep themselves and their partners off of certain lists that point to poor performance and market inefficiencies. Even if the specific culprit of a market crash in a symbol is not identified in real-time or end of day, someone else along the value chain that handled the order can be. Spur self-regulation of all market participants along the value chain by maintaining hourly or daily lists of those involved in market inefficiencies as measured by the market stability measures (circuit breakers) described above.  
Imagine the evening news and financial sites reporting not just the performance of the
daily market indices like the Dow and S&P but also showing the list of market inefficiency contributors. Everyone from the exchanges down through the chain of other identified participants will work to get his or her organization off the daily list.  They will also lean on and change behaviors of the participants, named or unnamed, that contribute to the inefficiency. 
 And here is Mark Cuban's point:
There is value to trading automation. It is here to stay. There is absolutely NO VALUE to High Frequency Trading. None. We need to bring our markets back to their original goals of creating capital for business.  It’s impossible to guess how many small to medium size companies have been held back from growing and creating jobs and wealth because of lack of access to capital from the stock market. It’s not impossible to know that our economy has suffered because Wall Street equity markets are no longer a source of equity for helping companies grow, it is not a platform for hackers and that needs to change. Quickly.

Full text : http://www.businessinsider.com/what-business-is-wall-street-in--2012-9

Monday, September 24, 2012

How to Hire the Best Computer Traders, And Fire Yourself

The Future of Investing and Trading is Here Now

Retail Investing is undergoing an evolutionary metamorphosis. I was watching the future in real-time while participating in a conference call meeting. On one hand, I was logged into the weekly Trade Ideas development meeting. There is never an end to our development, and our technology changes weekly to adjust to a market environment that is very different from the one that your parents invested in. Our rapid development, however, is not really the point of this blog post.

While I was speaking with development and discussing new feature priorities, my Trade Ideas software was trading the market for me, all by itself. Allow me to be more specific. The software was running strategies, 10 different strategies to be exact. Once again let me be even more specific. Each strategy inside Trade Ideas is like a trader that is specialized to look for a specific set up. One trader is looking for trading opportunities when the market is up and continuing to trade higher, one trader is looking to trade a sideways market, another trader is looking for downward pressure, and so on and so on.

This is not a black box system. This is the ultimate expression of technology helping the human to be better. Each of the strategies is relatively simple, to an extent. And, they're nothing like the sub-second arbitrage hunter robots that are much maligned by the press writing about HFT (High Frequency Trading). Each strategy's hold time may vary, some are 30 minutes, some hold times are hours, and some hold times are all day. A few of my computer traders are swing traders looking at the longer term trends and evaluating fundamental data along with price data. The bottom line is that the computer is helping me do something that I could not do even if I was 19 years old and hopped up on Aderol :-). My computer traders are dissecting information information faster, and with more accuracy than any of my human counter parts.
    • Tracking Social Activity? Check
    • Checking Multiple Time Frames? Check
    • Monitoring Profitability? Check
    • Scanning Multiple Technical Indicators? Check
    • Watching the stocks versus the whole market? Check and Check
As an investor or trader, unless you know something that everyone else does not know, this is the best you can do. That brings me to the next point. I learned to trade the hard way, via bloody noses. What does that mean? Well it means that as I traded, I made mistakes and applied lessons from those mistakes to my next trade. While I was one of the lucky few that made it, this was a very inefficient method. I made an immense amount of mistakes. My computer traders are way better than I ever was. Strategies don't get tired and they learn like a machine (machine learning). My computer traders play thousands of trades per day to find the most statistically viable options for the next day, and this continues day in and day out.  It looks something like this:

Calender Heat Map that is used by my Computer Trader

First notice that we don't win every day, that is ridiculous to assume. We do win a lot though. :-)

The best part of this was my realization of what was actually happening. Here I was in the middle of business that did not relate to actual trading. I was in the middle of that software development meeting and my computer was sitting there just taking the market to task, and making money! I had chills up my spine because I know for a fact that what I was witnessing is what every single investor wants. You may not be aware of it, but once you see it and taste it you can't go back. It's like swallowing the blue pill in the Matrix. It will change you!

Because of this experience, we started a whole new business called Execution Consulting. For those who want to experience it, we will help them create their robotic assistants to go out and crunch numbers and trade for them. So far what we hear from our customers is WOW!