Your Guide to Private Money Managers - Money Manager Review
Money Manager Review 
For a Free Tour of our Manager Performance Rankings
Call 1-415-386-7111
Follow Us Facebook Twitter
 
   Interview

    Guest Interview:

   System Research LLC

    21 Parkview Court
    White Plains,NY 10603

    Telephone: 914-582-5693
    Fax: 866-427-6393
    E-mail: info@sysre.com

 

    Interview Quarter: 2Q2011

 Vinay Munikoti

 CEO & Managing Member

  • Develops and Licenses quantitative program signals to Investment Management Firms. Firms currently allocate about $300 Million to System Research programs.
• Specializes in creating models that systematically combine technical and fundamental.
Prior Experience
•Partner
CL Capital Markets / Scalae Group LLC 2005-2008,
•Senior Quantitative Analyst
Glenwood Capital Investments/Man Investments 2002-2005,
•Quantitative Research Analyst
Rotella Capital Management 1997-2002.
Education
•M.Sc. Financial Engineering – Polytechnic University (Polytechnic Institute of NYU), May 1997
•B.Sc. Mechanical Engineering – University of Ottawa,
Canada, December 1994
 
  Vinay, when you analyze investments, which economic and technical indicators do you use?  
  System Research offers investment programs that trade the US Treasury Bond and the US Dollar. Both models systematically incorporate a number of technical and economic inputs. The Treasury Bond model’s technical inputs consist of momentum, moving averages and price channels while the economic inputs model the interest rate’s response to inflationary indicators as well as investors’ risk appetite/aversion. The US Dollar model’s technical input uses trend indexes created by combining multiple indicators over multiple timeframes. The indicators used are momentum, moving averages, breakouts, volatility bands and a swing indicator. The US Dollar model’s fundamental inputs model the relationship between the Dollar and market dynamics in hard assets, as well as the relationship between the Dollar and US/Global interest rate differentials which works as a supply and demand indicator.  
  What is your investment time horizon and how would you describe your overall investment strategy?  
  Unlike buy and hold or strategic asset allocation models, the programs are active and oriented more towards trading and attempt to capture market movements dynamically. The time horizon for trades is on average less than four weeks. The programs can be classified as ‘alternative’ or absolute return strategies as they provide long and short (positive and inverse) exposure to the US Treasury Bonds and the US Dollar. The investment programs are unique in that they are not driven exclusively by technical systems or by economic forecasting models, but a systematic combination of both.  
  What is your educational background and how much of your learning is based on broad economics versus quantitative analysis?  
  I have a B.Sc. in Mechanical Engineering from the University Ottawa and an M.Sc. in Financial Engineering from Polytechnic Institute of NYU. My undergraduate degree provided me with the mathematical discipline to design and implement complex systems while my graduate degree provided me with more specific knowledge on quantitative and numerical methods used in finance. The investment programs feature sub-models that look at economic or fundamental drivers of the asset classes. These models have analogies in textbook economic theory but some of them are tailored to reflect “real-life” or empirical market relationships that I have observed over my years in this industry. As an example, a hypothesis known as the “International Fisher Effect” states that a country with lower nominal interest rates will see its currency strengthen relative to a country with higher nominal interest rates because of the lower relative rate of inflation. However, over a short time horizon the opposite tends to occur. When investors are in a risk seeking mode, they will short low yielding currencies and use the proceeds to purchase higher yielding inflationary currencies in order to capture the yield differential. With a sufficient volume of market participants, there will be a weakening of the lower yielding currency relative to the higher yielding one. This is an example of a market relationship that is not necessarily taught in economics theory but does hold some practical merit.  
  What experience have you had as an investment manager, what did you learn, and why are you now on your own?  
  Prior to starting System Research I was employed as a quantitative analyst at 3 different firms, starting in 1997. Two of those were managed futures funds and one was a fund of hedge funds. At the managed futures firms I was responsible for developing trend/technical models across all asset classes and was in a position to provide input on the merits of implementing my strategies directly to the portfolio manager. At the fund of hedge funds, part of my job was to provide inputs to the investment committee on including certain strategies in the portfolio based on my quantitative research and qualitative assessments; I was involved in assessing a range of hedge fund strategies including CTA, global macro and statistical arbitrage/market neutral. After more than a decade of working at these firms, I felt that it was time to go on my own and draw upon all my experience in order to design and offer clients unique programs that systematically combine price based models and fundamental/economic forecasting models. I have been running System Research since 2008 and have experienced a steady growth in client base and AUM.  
  What are the current Assets Under Management in System Research programs?  
  Currently there is approximately $300 Million of assets traded utilizing System Research programs. System Research licenses its investment programs to a range of investment management firms that manage their assets using the program signals. Client firms include SEC and state registered RIA firms, regional broker/dealers, hedge funds and mutual funds. Additionally, I manage client assets directly through separately managed accounts.  
  Why do you frequently adjust your investment positions and what is the typical turnover rate of a portfolio?  
  The programs are characterized by the active switching between long and inverse positions. This is the nature of the programs as they are constantly looking at trailing market data in order to update the technical and fundamental directional forecasts. The portfolio turnover rate/ trading speed varies by program. The TBond-LT program does about 30 trades per year with an average trade length of approximately 12 days. The Macro-LT program does about 50 trades per year with an average trade length of approximately 16 days.  
  How important to your investment is the ratio between market sectors? Do you ever go 100% in a single market sector?  
  The multi asset program Macro-LT trades the US Treasury Bonds and US Dollar. This program does not look at the ratio or relative performance between T-Bonds and Dollar in order to make investment decisions. Instead the T-Bond and Dollar models are traded independently and each one is weighted such that they each contribute an equal volatility to the overall program. The purpose of combining the T-Bond and Dollar models is to provide the client with the option of investing in a lower volatility diversified multi-asset program.  
  Do you ever use market leverage, or would you consider it?  
  I do not use leverage through the use of margin. Some of the instruments I trade have a built-in leverage. Rydex offers +1.20/-1.00 Beta to Treasury Bonds and +2.00/-2.00 Beta to US Dollar. Profunds offers +1.25/-1.25 Beta to Treasury Bonds and +1.00/-1.00 Beta to US Dollar. In the ETF programs, I use unleveraged as well as leveraged exposure ETFs. The investment programs take into account the built-in leverage of the traded instruments as well as the changing volatilities of the asset classes in order to calculate the daily allocations in the portfolio. Also, some investors are concerned about the tracking error in leveraged funds and ETFs. In my programs that is not an issue because of the relatively active trading in and out of the securities.  
  I assume you are not an adherent to the random walk theory of the market. What is it about your approach that gives your system a better than 50% market edge?  
  That’s correct, I do not believe that markets are totally random and I do believe that markets often move in response to certain driving factors. I believe that well constructed investment programs can take advantage of these cause and effect relationships and beat buy and hold consistently over the long term. Some managers implement their active strategies using trend or momentum analysis while others invest based on their assessment of the fundamentals and economic drivers. I believe what gives my programs an edge is that they use a combination of multiple sub models that incorporate non-correlated technical and economic inputs. As an example, the TBond-LT program is made up of 4 sub models where each one contains one or more economic and/or technical inputs. Each sub model has been designed to have low correlation to the other sub models and the average pair-wise correlation among the sub models is near zero. So you could say that my programs aim to combine multiple edges that are not correlated to each other.  
  Do you ever short the market and what triggers the signal?  
  The programs do not have a directional bias and can be long or short with equal probability as the underlying models have symmetric “go long” and “go short” rules. When the programs signal to go short, I do not actually sell short. Instead I buy the inverse exposure fund. This enables the programs to be traded in taxable accounts as well as in tax deferred accounts where there might be restrictions on shorting.  
  What kind of investment instruments do you use and in what combination?  
  Most of the clients participate in the TBond-LT and Macro-LT programs through the Rydex long and inverse exposure funds. The TBond-LT program trades the Rydex Gov’t Long Bond 1.2X Strategy and the Rydex Inverse Gov’t Long Bond Strategy. The Macro-LT program, in addition to the two bond funds mentioned, trades the Rydex Strengthening Dollar 2X Strategy and the Rydex Weakening Dollar 2X Strategy.  
  Some say that quantitative market analysis has a bad reputation but isn’t it a fact we now have more computers driving more investment decisions than ever?  
  Quantitative analysis and quantitative investment models can come in several different forms. In the past there has been some notoriety associated with ‘high frequency’ trading programs. These programs constantly scan across thousands of securities or futures contracts and trade in and out of the market with holding periods of just a few seconds of less. By necessity, these programs are run autonomously by computers with little human intervention. On May 6th 2010 a large short trade in the futures market initiated a cascade of events that eventually caused the stock market to lose about 7% of its value over a 15 minute time span and then spike back up to partially recover some of the losses. This market action is generally attributed to high frequency trading programs whose algorithms were ill equipped to handle certain market dynamics and that resulted in a feedback effect among competing programs. My programs, while classified as quantitative, have no relationship to the above. The term ‘quantitative’ comes from the fact that I process the raw market data through my computer based models in order to generate buy and sell decisions. Once the models are run and have generated their buy/sell signals, I manually input trades into the managed account platforms to transact at the market close.  
  If trend analysis made everyone money we would all be rich. What is different in the way you use it?  
  Trend following systems do make money over the long term but the problem is that most trend models are characterized by prolonged periods of drawdown, during which it is psychologically very difficult to stay invested and most people tend to stop trading near the bottom of a drawdown. To lessen this effect many trend traders try to diversify by using multiple trend models by varying the technical indicators and the lookback periods. This does provide some staggered entry and exit points but the correlations among the systems will still be high, probably greater than 40%. The difference in what I am doing lies in the fact that my sub models contain fundamental and/or technical inputs and they have been designed to each provide unique information to the overall program and thus avoid information overlap - the average correlation among the sub models is near zero. This makes for a more stable program since the sub models are not moving in tandem and therefore there is a reduced probability of everything being wrong for a prolonged period.  
  Are your accounts set up to automatically trade when signaled by your model?  
  The signals are generated by computer models but the accounts are not traded on “auto-pilot”. Each day, raw data is imported into the models and processed in order to generate the trading signals. Once the signals are generated I log into the various trading platforms and manually input the trades at the managed account level and those trades are automatically distributed across the client accounts.  
  I look at your performance and say “Wow, that’s a great investment record”. Now prove to me that the numbers you are showing me are real.  
  Sure. My programs are traded at Rydex and Profunds accounts that are verified daily by Theta Investment Research and recorded in their database. Theta Investment Research is a firm that has been providing track record verification services for money managers since 1999. My programs are also verified at TimerTrac, which is another performance verification firm that does daily tracking of my programs. My track records posted on Money Manager Review are from the verified performance from Theta Investment Research.  
  Please tell us the difference between System Research TBond-LT and Macro-LT programs. And for the last several years, why has TBond-LT done so much better than Macro-LT?  
  TBond-LT trades the Treasury Bonds whereas Macro-LT is a blended program trading 50% Treasury Bonds and 50% US Dollar. The US Dollar program is a stable and highly developed program just like T-Bond program but it does have an expected return of about half that of the T-Bond program over the long term. Then the question arises “why combine the T-Bond program with a lower performing program such as the Dollar program?”. The answer is that the programs have zero correlation to each other and the combination of the two produces higher risk adjusted returns than either program alone as measured by Return/Volatility and Return/Drawdown. So the combination program Macro-LT is appropriate for clients who are willing to accept more subdued performance in exchange for enhanced risk adjusted parameters and lower volatility.  
  What is the most money you think you can handle before your model has liquidity problems?  
  When traded through Rydex or Profunds mutual funds I would say that the upper limit before liquidity becomes an issue is quite high. As an example, the total assets in the Rydex and Profunds Treasury Bond funds currently exceed $1Billion. These mutual fund operators are sophisticated players and they would be able to handle transactions of substantial size as they obtain their exposures through a variety of instruments including OTC derivatives. Also, if assets grew beyond a certain level, I would probably look for additional liquidity that is available in ETFs.  
  Are accounts established in the account-holder’s name?  
  Accounts are established and held in the client’s name and the client gives my investment advisory firm trading authority. Clients can participate in my programs through taxable as well as tax deferred accounts since I don’t use leverage or short selling. Short exposure is achieved by buying inverse funds.  
  What is your current account minimum and what is a realistic time period before I can assess the adequacy of your performance?  
  My minimum account size is currently $25,000 which is subject ot change. I would say that a client should stay invested in my programs for at least 36 months in order to capture the true nature of the programs and to participate in their return potential. Some clients do tend to jump in after a period of strong performance or pull out at the bottom of a drawdown in their cumulative return but that usually results in them not realizing what the programs can offer over the longer term. So I would say that clients would benefit by holding steady through the day to day volatility and by ‘buying and holding’ actively managed programs.  
  Can I assume your investment model is proprietary and do you ever tweak your model to see if it can be improved?  
  Yes the investment models are proprietary but the general themes and economic drivers of the programs are disclosed so it’s not a total “black box”. As a quantitative model based trader, my goal is to always have the most robust and stable models at any given time. This is achieved through ongoing research and by always seeking new sources of edge. The overall policy is to leave the models alone but an enhancement might be implemented on rare occasion if I identify an edge that is in my opinion unique, stable and real. A new edge can be in the form of an enhancement to the sub-model weighting mechanism, an enhancement in risk management or an additional predictive variable.  
  Many investor portfolios have significant exposure to the stock market. How would your programs fit in with those portfolios?  
  Due to the active long/short trading of non-equity instruments, both of the programs that I offer have low correlation to the stock market. Since inception, the correlation between the Macro-LT program and S&P500 is 0.10 and the correlation between the TBond-LT program and S&P500 is -0.01. This makes my programs a great diversifier of traditional long only and asset allocation/strategic portfolios. The lack of correlation was clearly seen when the programs made money during the stock market meltdown of 2008. Both programs had a profitable 2008 and have not had any losing years since inception (disclaimer: past performance is not necessarily indicative of future results).  
  Where can I get more information on System Research and how do I open an account?  
  There is additional information about the firm at www.sysre.com. Prospective clients can contact me by phone at 914.582.5693 or email me at info@sysre.com for further information on System Research, the investment programs and account opening information.  
 
 
 
 
Manager Services | Site Tour | Site Map | About Us | Interviews | Contact Us
© 2014 Money Manager Review TM. All Rights Reserved. Terms of Service | Disclosure & Privacy Statement1.