TABLE OF CONTENT
ProQuant lets anyone build and follow automated trading strategies for stocks, forex and cryptos via an easy to use mobile app. It combines complex algorithms, powerful cloud computing and simple user interface to make algorithmic trading accessible to anyone, not just the big Wall Str. players.
With the ProQuant app it takes only a minute to run your first trading strategy. It is perfectly enough to choose one of the existing strategies in the Strategy Library and press “Run” in order to start automatic trading. Or geek out and go as deep as you want by experimenting with various parameters to generate your own customised, tailor-made strategy.
This is made possible by the Strategy Generator, an innovation of ProQuant - using complex algorithms, it composes strategies automatically, evaluates each strategy by backtesting it against historical data, presents the results in stats and charts and shows the most profitable strategies found. Thus, instead of manually testing the performance of each combination of various rules, traders can set their own basic criteria and ProQuant generates and backtests thousands of possible strategies to present a list of the top performers. The whole process takes mere seconds, then the trader can immediately save the strategies, run or modify them. The PQ strategies run in the cloud utilizing vast computing power so that traders won’t have to deal with any technical issues.
Then you can run your strategies for as long as you’d wish in a risk-free real-market simulation before trading with real capital, which is another valuable option that ProQuant gives to its traders through an easy connection to trading platforms, MetaTrader 4 being the only available at the moment with more in the pipeline.
In ProQuant, strategies are sets of rules for opening positions (Entry Rules) and for closing positions (Exit Rules) which trigger BUY or SELL signals for a great number of financial instruments. Currently, the rules are based on combinations of standard and widely used technical indicators.
are indicators that determine the conditions for opening a position. A strategy can have up to four Entry Rules. ProQuant uses these rules to evaluate whether a position should be opened and in what direction. All Entry Rules are evaluated periodically at each bar, depending on the trading frequency of the strategy. This evaluation is done via backtesting, which is an automated testing of the proposed strategy against real-market conditions within a certain period of time in the past. The evaluation of each Entry Rule results in an “answer” or “vote”, which shows whether the strategy should open a position in a certain direction. However, we wanted ProQuant to open a position only if all Entry Rules requirements are met. In other words, if one Entry Rule produces a signal in a given direction, but another one doesn’t, ProQuant will not act and will not open a position at all.
are indicators that determine the conditions for closing a position. A strategy can have a maximum of two Exit Rules. Even when we have the maximum number of two Exit Rules, the strategy will close the position at Bar Open if one of them is satisfied, regardless of the other, which constitutes a major difference between the Entry Rules and the Exit Rules.
You can acquire strategies in two ways:
Go to the Library and select one of the ready-to-use Popular strategies there to save it for later use or run it immediately (all these strategies have done exceptionally well during the backtest)
Go to the Strategy Generator, configure and run it to generate, backtest and filter top-performing strategies. First, you must choose an instrument to your liking, be it stocks, forex, commodities, cryptos or indices. After an instrument has been selected, you will have to take a look at the three major properties of the strategies to be generated - Funds for Trading, Leverage and Trade Size - and either accept the default values or change them in the preferred direction.
At any point, while setting up the Generator, you have the ability to dive deep into its Advanced Mode in order to select and set preferred values to any number of the additional settings that are in stock, along with the so-called Acceptance Criteria. All you need to do is scroll down and choose from the list:
Trading frequency sets the frequency for the evaluation of trading signals, i.e. the time interval for each successive evaluation of the market conditions that enables the strategy to decide whether or not to open a position. The available time intervals are as follows: 1 minute, 5 minutes, 10 minutes, 15 minutes, 30 minutes, 1 hour, 4 hours or 1 day.
Max data bars is a filter which sets the number of data bars (10 000 - 200 000) that will be used for the backtesting. A bar is the representation of price movement for a set period of time. The more bars you have, the further back in time the backtest goes. Bear in mind that the available historical data for some instruments, like cryptos, is limited.
The Optimizer is a tool that will attempt to make the current strategy better. The Optimizer tries out different variants of the same strategy to get better results on the backtest. The Optimizer algorithm does not brute-force the parameters, thus avoiding curve-fitting and allowing for the creation of strategies that are closer to real-life market conditions.
Monte Carlo is the best tool for testing the viability of a strategy. When a strategy is generated, there is a chance for it to be over-optimised (curve-fitted), which may happen if the real-market situation in the past period of time, covered by the backtest, featured some extraordinarily favourable conditions that are unlikely to occur again. Monte Carlo applies random changes to the market data, to the execution of the strategy and to the numeric parameters of the strategy indicators, in order to introduce some “noise” and make sure that the strategy is not over-optimised. If Monte Carlo makes minor changes to the strategy and its environment, and the strategy still comes profitable enough after the backtesting, then it has a good chance to make money in the real market as well. On the other hand, if the strategy profits crumble after the minor changes, effected by the Monte Carlo tool, this would mark it as over-optimised and a bad choice for trading.
If selected, this filter will tell the Generator to evaluate only long entry/exit conditions.
In order to evaluate the balance of a strategy, some kind of reference must be used, like an imaginary Reference Line, which shows how the strategy balance should move were the conditions perfect. This best-case scenario is highly improbable, of course. The actual balance line will most certainly deviate from the Reference Line. The Maximum Balance Deviation criterion (MBD) sets limits on how far the real balance is “allowed” to move away from the Reference Line. For example, if the MBD has been given the value of 20%, no point along the actual balance line should stray with more than 20% (up or down) from the Reference Line. MBD is a measure of the volatility of the instrument during the backtested period of time.
Sets the maximum consecutive losses allowed per strategy during its backtest. The Generator will only show you strategies with less consecutive losses than the number you have specified. Very low settings for this criterion might be considered restrictive, especially if a longer backtest period has been opted for.
Sets the maximum number of trades the strategy is to make in the backtest.
This criterion defines the maximum allowed drop from peak to trough before the next peak during the backtest, described in percentage from the highest peak, e.g. with a $2000 peak and 20% drawdown, the maximum allowed difference will be $400. The maximum drawdown metric needs to be in line with the trader's risk tolerance and the trading account size - anyone can win a million, if they have and are willing to risk ten. In other words, if the largest amount of money that a trader is willing to risk is less than the maximum drawdown, the trading system is not suitable for them. Bear in mind that this criterion is time-dependant, i.e. the longer the backtested period is, the higher the maximum drawdown of the tested instrument is likely to be.
Sets the maximum number of consecutive days in which the strategy is not making profit, described as a percentage of the full backtest duration.
Sets the minimum average Holding Period Return in percent. In finance, Holding Period Return (HPR) is the return on an asset over the whole period during which it was held. When measured, the difference between the ending and initial values of the investment (V2 – V1) is divided by the initial value of the investment (V1), hence the percentage. It is one of the simplest and most important measures of investment performance.
It is set to 100 by default. This means the strategy should make at least 100 trades during the backtest for the results to be considered trustworthy. If there are too few trades, the results from the backtesting may look good, and yet harm your profit, if you decide to run the strategy. For example, you may have a good win/loss ratio, but if only one trade has been analysed in the backtest, then this strategy is not reliable.
When tested, the strategy should make at least the specified amount of profit, in your account currency. Bear in mind that if you set the value for this criterion to, let’s say, $5000, the Generator will ignore all strategies with net profit of $4999, $4998 and lower.
Sets the minimum profit factor of the strategy. Profit factor is a measurement of one’s trading skills with respect to the amount of risk they are willing to take. In our case, the Generator takes into account the gross profits of the strategy, dividing them by its gross losses, to find out how well the strategy is doing. Or not so well, as the case may be, if its profit factor is lower than 2. Values greater than one indicatе a profitable system; profit factor of 1 means that the investment has returned a zero profit while taking the risk of losing the initial equity; profit factor below 1 indicates a loss. Very high PFs are probably the result of short trading periods.
Sets the minimum amount of profit strategies are required to make per day, in your account currency.
Sets the minimum profit/drawdown ratio that the tested strategy is allowed to use, where the return to drawdown ratio is the average return expressed as a proportion of the maximum drawdown level. This ratio of the average return on an asset against its drawdown gives a fairly good idea about the risk involved. Although drawdowns are only natural to trading, investors usually look for assets that have had a return about twice as big as their maximum drawdown.That means that if the maximum drawdown has been 10% over a given period, investors will want to see a return of 20% (the return/drawdown ratio, known in finance as RoMaD, would be 2 in this case). This, literally speaking, means that each time you hope to earn 2 dollars, you can expect to lose maximum 1 dollar per the period in your average worst-case scenario. Or, to simplify it further, an investor should ask themselves, am I willing to accept an occasional drawdown of X % in order to generate an average return of Y%? If you set this criterion to a high value, the Generator will only backtest strategies that have been relatively conservative with respect to their RoMaD (low drawdown and high return), which might prove restrictive.
Sets the minimum Sharpe ratio the strategy is allowed to have. This is yet another way to try and adjust risk to return. As the theory goes, the higher the Sharpe ratio of an asset is, the better its risk-adjusted performance is expected to be. In simple terms, it shows how much additional return an investor earns by taking additional risk. In other words, if portfolio X generates a 10% return with a 1.25 Sharpe ratio and portfolio Y also generates a 10% return but with a 1.00 Sharpe ratio, then X is the better portfolio because it achieves the same return with less risk. This criterion is highly restrictive when set at a high value. Generally, a Sharpe ratio of 1 or better is considered a good risk-adjusted return, 2 or better is very good, and 3 or better is excellent.
The System Quality Number (SQN) can be interpreted as the overall trading “grade”of a strategy. It helps building position sizing strategies that are likely to achieve one’s return objective, while avoiding undesirable drawdowns. Position sizing sets the specific amount of funds an investor is willing to risk with every trade (see Trade Size above) and is in direct connection to the total amount of the account (see Funds for Trading) and Stop Loss. In theory, a SQN of 1.6 to 1.9 is considered to be below average but tradable, while a SQN of 5.1 to 6.9 is superb. Bear in mind that a backtest with a very high number of trades can generate a very high SQN number.
The Win/Loss Ratio is calculated as follows: (number of profitable trades) / (number of profitable trades + number of losing trades) Its value varies between zero and one. The win/loss ratio is the ratio of the total number of winning trades to the number of losing trades. It does not take into account how much was won or lost, but simply if the trades were winners or losers. Thus, you may have a good win/loss ratio and still be on the losing side, if your good trades have returned less money than your bad ones have lost. If you select the value of 1 for this criterion, you tell the Generator to select only strategies with zero losing trades.
The Generator composes sets of entry/exit rules (i.e. strategies) and simultaneously filters them down according to the specified acceptance criteria. This whole process happens so fast that the Generator can evaluate hundreds of strategies per second.
Backtesting assesses the viability of a trading strategy by discovering how it would play out with historical data. It allows a trader to simulate a trading strategy using historical data to generate results and analyse risk and profitability before risking any actual capital. A well-conducted backtest that yields positive results is a sign that the strategy is fundamentally sound and is likely to yield profits when implemented in reality.
ProQuant evaluates the entry/exit rules bar by bar, from the beginning to the end of the historical data for the chosen test period, and calculates the trading results. When the backtesting is finished, the Generator shows you the top performers from the virtual trading. If you get very few results, or none at all, this is a sign that some of the criteria settings you have specified are too restrictive.
You can additionally edit, your saved and running strategies, by introducing Stop Loss and Take Profit orders or changing their settings if the strategy already has them :
Stop Loss is an order to sell a financial instrument when its price hits a prespecified low point. The idea behind the Stop Loss orders is to limit an investor’s loss.
If the Take Profit tool is activated, ProQuant automatically closes the position once the specified profit has been made, which could preclude a further gain, but is a hedge against possible loss, should the price of the asset suffer a nose-dive.
And this is how we get to greatest advantage of using ProQuant. Our app enables you, by choice, to switch to real-market trading via the MetaTrader4 platform, with MT4 being one of the many options, although the only one currently available, for automated trading (more are on the way). Connecting with MT4 will let you execute ProQuant signals automatically in your MT4 broker. If you decide to opt-in, an email will be sent to you containing a bridge that you will have to import in your MT4 platform. You may enable any strategy to trade automatically in MT4 at any point, even while the strategy is running, as long as you've requested and imported our MT4 bridge correctly.
After you have successfully tailored a strategy and selected an instrument, the actual trading begins. In implementing it, ProQuant follows some strict trading rules:
All positions open at the beginning of a bar. We use the term Bar Open to identify that. Bar Open is the first tick of a new bar, where tick is the first infinitesimal change in price for a period of time.
The strategies are symmetrical – the long trading rules are exactly opposite to the short trading rules. This principle guarantees that the strategies will work on both bullish and bearish markets. If they spot a pattern, they will act on it.
The strategy will open a position when all the Entry Rules are satisfied.
If all available Entry Rules are satisfied in both directions, this means there is no “clear” signal in which direction to trade, therefore the strategy will abstain from opening a position.
The strategy will close its position at Bar Open. Exceptions are the Stop Loss and the Take Profit tools.
The strategy will close its position even if only one Exit Rule is satisfied. Similar to the Entry Rules, the Exit Rules are symmetrical for long and for short positions.