A Statistical Approach
The concept behind Novofina is based on precise and purely statistical methods. We examine the recurring effects and regularities of stock prices on the principle of mean reversion (reversion back to the average) and operate each system strictly according to this pattern. Our systems look out for irregular short-term deviations from the mean that occur within a few days. In terms of statistics, there is a higher probability that the price will revert back to the mean following such an exaggerated deviation from it. We use this probability shift to generate long-term positive returns.
The Principle of Mean Reversion
“The higher the climb, the harder the fall”
The principle of mean reversion has been traded successfully for many years. The Nobel Laureate and financial entrepreneur Robert Shiller has preferred this approach over many other methods.
Shiller received the 2013 Nobel Prize in Economics for the discovery of this effect and the co-founding of the underlying behavioral finance theory.
The mean reversion theory describes a recurring pattern, it shows how exaggerations in stock prices tend to correct themselves over time back towards the mean. Even in bull markets, the path of the stock price would not be rising continuously day after day. One can normally witness that a few days of repeated setbacks in the opposite direction would have occurred regularly.
Thus, we experience phases of impairment and improvement in the price of each security, which is similar to the motion of a wave. This occurs in the short-term and over a few days. If such price fluctuations are particularly intense, the probability of reverting back to the average within the next few days increases. These are the patterns analyzed by our systems for a broad selection of US stocks on a daily basis.
The phenomenon of mean reversion is independent of the direction of the stock price trend, and can be observed even in falling or stagnating markets. The effect occurs consistently both in times of economic growth, as well as in times of crisis.
A Statistical Advantage and the Law of Large Numbers lead to Success
At Novofina we rely on the recurrence of these corrections following deviations from the mean price – over and over again. Sometimes we have a successful outcome while at other times, we do not. Our systems cannot predict whether each and every individual trade will be successfully or not. But we know that on the whole we are more often right than wrong:
In approximately 55 to 60 percent of all cases we record a winning trade and in about 45 to 40 percent of cases a losing trade – empirically verified.
The net proceeds of winning trades (= the amount credited to your account after fees have been deducted) are also slightly higher than the net losses of losing trades.
The combined result of both of these positive effects is sufficient for us to be profitable over time and to minimize and manage the risk of loss and temporary drawdowns.
This statistical advantage that we have achieved through our approach can be leveraged only by a high volume and frequency of trades. For this reason, we use the law of large numbers: A principle of probability and statistics which states that as a sample size grows, its mean will get closer and closer to the average of the whole population. In order to achieve the necessary critical mass, we use highly specialized trading systems which analyze a variety of stock prices according to this existing pattern. These rules are reproducible over short intervals and thus, our investments are made with a holding period of only a few days. The breakdown of the investment capital across many small trades result into an associated high level of diversification which in turn reduces the overall risk.
An enormous advantage results due to the independence from the direction of stock trends and market movements. Because opposed to traditional equity funds, we can make a profit in all market conditions, even in falling and stagnating markets. We can expect high stability and reliability of our system to achieve consistently positive returns even in times of crisis. We do not rely on inaccurate medium to long term market forecasts. This also helps to increase the predictability of your system.
The bottom line is simple: Real profits in your account attributable to diligently implemented financial science. Independent from market directions, year after year.
Double Protection With Backtesting
This Is How We Review Our Performance on the Real Market.
A great advantage of trading systems is that all trading rules (entries, exits, position sizing) are clearly defined. Moreover, each strategy’s expected performance can be measured through back-testing with real market data.
During the back-testing phase, a trading system with all its specifications is applied to real past security prices over any period of time (e.g. 10 years) and analysed. Through this exercise, the expected return that would have accrued had the strategy been used over the particular period can be determined. Furthermore, back-testing also allows us to calculate how large the maximum setback in the capital curve (i.e. drawdown) would have been and how the strategies would have performed in crisis years, such as the dot-com bubble in 2001 and the banking crisis of 2008. At Novofina we are committed to test and optimize our systems regularly based on this principle in terms of their performance (return) and stability (risk).
Although back-testing results have a very high general significance: Past performance is no guarantee of future performance, thus the performance of our trading systems can vary to some degree in both directions from the results produced by our back-tests. To account for this uncertainty, we always adopt a conservative estimate of the expected average annual return of our products.
All Risks Considered: The Helnwein-Expectation (HE) Value of Your Return
Our yield estimation measure, the Helnwein-Expectation (HE) value, is based on the back-testing results of the last 10 years. The returns achieved over the testing period are not simply averaged, but are subject to reducing expectations calculations. We deliberately take a very conservative measure of the expected return on our products.
Step 1: Positive Outliers are Excluded
In the first step of the calculation, the returns produced in the best two years are treated as outliers and therefore not taken into account from the outset. For the other years, we take note of the average value.
Step 2: Creating a Forecast Corridor
Next an expectation corridor is formed by the standard deviation which is calculated and plotted against the already lowered average taken in step 1. The standard deviation is a common measure of risk and it takes into consideration the deviations from an average value. The remaining particularly positive outliers are therefore also excluded.
Step 3: Interpretation of the Forecast Corridor
To determine the Helnwein-Expectancy value, we focus only on the lower limit of the forecast corridor. This means that the performance values specified by us always refer to the lowest performance of our forecast corridor – in this example, 7% of the possible 21% expected return. So we end up with a conservative expected return, low risk and high growth potential.