Frequently Asked Questions

Contents

  1. What is the underpinning investment philosophy of Qantares?
  2. How does the investment strategy work?
  3. Are there constraints for QanSystem?
  4. What are the strengths and weaknesses of QanSystem?
  5. What is the competitive advantage of QanSystem?
  6. What do your IT systems offer in the way of failover to ensure trades can be executed in a timely way?
  1. What is the underpinning investment philosophy of Qantares?

We believe that none of the models known and applied in finance adequately depict reality. We therefore use non-parametric techniques, with no model or any other assumptions in order to forecast the future and to price options. This is a mathematically intricate task and, to our knowledge, has not previously been achieved by anyone else. With our approach, we are able to capture not only regular market behaviour more realistically but also to uncover very unfamiliar patterns. For example, a highly asymmetric distribution could reflect a strong directional bias, or it may suggest limited market potential in one direction as opposed to possibly substantial movements in the opposite direction.

Further, it may even be possible to assess the likelihood of a crash or to identify rare situations such as before an imminent and unpredictable interest rate decision, which are best described by a bimodal distribution of future market behaviour.

Every forecast is made with a greater or lesser degree of certainty. To qualify the accuracy of a forecast and to provide for professional scrutiny, with every forecast QanSystem produces a measure of reliability indicating how much faith one can have in the forecast. This is an ideal gauge for determining position size. Of course, it also implies that there may be periods of time with limited trading activity.

Over the 15 years since 2002 and with respect to a 5-trading-day forecast horizon we determined an average signal-to-noise ratio amounting to 29% signal and 71% noise in the DAX time series. In that sense, part of the future is embedded in the past, which we are able to extract using state-of-the-art mathematical and statistical techniques. Although there are times when this signal is weak or when the QanSystem forecasts do not differ much from common market assessment, usually this is not the case and arbitrage opportunities arise.

Finally, when it comes to trading markets, there are two principal camps: the range traders, always adopting a contrarian view, and the more dominant camp of trend followers. While the former group achieves good results in sideways markets, they usually perform poorly in trending markets. For the trend followers it is obviously the opposite. By contrast, QanSystem, which is based on a self-learning algorithm, strikes a fine balance between trend following and range trading, seeking opportunities both ways and, at the same time, eliminating the pitfalls of other systems.
Back to top.

  1. How does the investment strategy work?

The short-term distributional forecast contains expected market direction and other characteristics. Using a set of pre-defined trading rules all these characteristics are translated into trade instructions. It follows that the market has to be observed continuously to ensure that market opportunities are seized and, even more importantly, that positions are closed when trading signals reverse and markets move in the opposite direction to what had been expected when the position was opened.

As long as position or risk limits are not violated, orders are placed on the basis of these trade instructions. Both distributional forecasts and trading signals are calculated every 15 minutes in real time in QanSystem. The trade instructions are then converted into orders which are routed through to the appropriate trading platform such as Eurex for DAX futures and options. QanSystem is built such that it can be used as an automated trading agent.

The investment decisions are solely based on the instructions generated mathematically by the QanSystem execution engine. Additional market research or analysis is not required to operate the fund.
Back to top.

  1. Are there constraints for QanSystem?

Since the DAX futures market is very liquid, there are no capacity constraints for practical purposes. This situation may be different when trading other markets where liquidity is insufficient. As with all trading models, the primary constraint is liquidity of both the underlying and traded instrument markets. For QanSystem low liquidity generates two issues. First, the less liquid the underlying market, the less reliable the distribution forecast and therefore the quality of the trade indications. Second, trade size is limited to positions that can be quickly opened or closed without moving the price and reducing profit.
Back to top.

  1. What are the strengths and weaknesses of QanSystem?

Strengths:

  • No assumptions are made for QanSystem thus avoiding model rigidities and difficulties in capturing true market behaviour. As a result the strong performance is independent of the prevailing market situation.
  • As entire distribution functions are forecasted, QanSystem generates not only directional indicators but a wide range of additional characteristics such as:
    • Extent of movement measures,
    • Confidence measures,
    • ‘Sharpe-ratio’ like risk / return measures built around expected profit, expected loss, expected volatility, expected profit conditional on the forecasted market direction being right or wrong, etc., and
    • Significant movement indicators.
  • This is possible because a distribution function conveys the complete picture of stochastic variability, leaving little scope for unwelcome surprises.
  • QanSystem demonstrates outstanding forecast performance.
  • The QanSystem approach can also be used to price options, both to enhance profit capture and to mitigate / manage risk. It is in this context where the greatest arbitrage opportunities can be exploited.

Weaknesses:

  • As with all trading models, there may be periods when the implied strategies do not work and losses occur. However, over the years since 2002 these periods have been short-lived and, except on the rarest of occasions, have been associated with measures of reduced reliability (implying that only small positions, if any, were affected).
  • There may also be periods of time with limited trading activity, as there are situations when directional indicators are balanced and / or directional indicators have low confidence statistics, i.e. only weak or incomplete trade signals are obtained.
  • As the self-learning algorithm is fundamental to QanSystem, problems can occur if an established stable pattern becomes invalidated because of a sudden and significant market regime change. This occurred in June 2006. In this case QanSystem will re-learn the new market conditions and this may take a period of up to four to six weeks to generate reliable forecasts again. During such times, no trade recommendations are made, preserving investor capital.
  • Many process driven trading models generate ‘over-trading’, i.e. significant, sustained and expensive executions with little profit. By using confidence of profit as a key indicator of trading, the set of rules and trading discipline will act as a natural limit on this practice for trades generated by QanSystem.

Back to top.

  1. What is the competitive advantage of QanSystem?

There simply are no competitive offerings worldwide today which are based on a similar mathematical-statistical foundation as QanSystem. To our knowledge, no one else has ever devised a mathematical concept of a non-parametric stochastic process which can be used to reliably and accurately capture financial market behaviour.

In a nutshell, QanSystem:

  • Provides forecasts of the entire distribution function,
  • Is free of any assumptions,
  • Delivers a reliability measure indicating how reliable the forecast is, and
  • Reaches a success rate that is close to what is conceptually possible to achieve.

Back to top.

  1. What do our IT systems offer in the way of failover to ensure trades can be executed in a timely way?

Client financial outcomes when using QanSystem are of the utmost importance to Qantares. With demanding global clients, the compute-intensive and data-intensive operations imbedded into our trading system are met by a high performance analytic server farm supporting the most time-critical decision processing. Not only do we have a fully-encrypted and highly secure trading gateway, our underlying IT systems operate in a fault-tolerant fashion, where resiliency and redundancy is built in. Our geographically dispersed systems ensure the highest availability of QanSystem so that trade instructions can be calculated instantaneously at any time and executed according to client trading strategies.
Back to top.