ALGOLAB PERFORMANCE VIEWER
Using the Performance Viewer backtesting application with capital settings under $100,000: if you have less than $80,000 in your brokerage account, AlgoLab will automatically reduce the number of symbols traded to an optimized set of symbols more appropriate for your capital amount. Performance Viewer does NOT automatically reduce the symbols. So, a backtest using PV, with capital less than $80,000 will result in estimates of performance that use the FULL symbol set. The table below shows performance estimates for each trading system and various capital levels, along with the resulting number of symbols traded for each in AlgoLab.
To opt-out of the automatic symbol reduction, set your "capital" field in your AlgoLab dashboard to an amount above "$110,000".
The AlgoLab Performance Viewer 2 web application here: dashboard.thealgolab.com Select the "Performance Viewer" tab on any dashboard page.
View historical performance metrics for all AlgoLab's included MultiSystems
Adjust your capital and leverage settings and learn how that effects your returns and drawdowns
Set custom date ranges to inspect performance results over a specific date period
Customize the symbol portfolio
Use Day of Week, Month of Year and Lunar Phase filters to learn how they effect performance metrics
Trading Systems Performance Expectations
Algorithmic trading systems have been developed by optimizing system rules, and parameters on historical data. This does not guarantee that past performance will be as profitable, or will exhibit similar characteristics to real-time trading results. Given enough variables, it is possible to curve fit a system to historical data, and any relationship between those rules and future, unseen data may be random. AlgoLab has taken steps to reduce the number of variables and system rules to reduce the degrees of freedom which will reduce the chance of curve fitting. Current market regimes can and do change, and rules that previously predicted future prices may no longer work.
About the AlgoLab Performance Viewer
Use the AlgoLab Performance Viewer 2 to learn more about the profit and loss characteristics of an AlgoLab trading system before you start trading it. The Performance Viewer 2 is a html 5 web application that runs in your browser window. You can view historical backtesting results from 2009 to the most recent update. With PV2, you can specify the amount of capital you have to trade with, and how much leverage you are willing to use. Note that a higher leverage setting will usually increase your return, but will also increase your drawdown (losses). You can also experiment with changing the portfolio of symbols that your system trades and view the performance metrics resulting from those changes. You can inspect results by going long or short only, and you can also change the date range to see detailed performance results over a smaller period of time.
Overview of The AlgoLab Performance Viewer 2 Window
TRADING SYSTEM Select a trading system to view historical backtesting performance results
SYMBOLS Specify which symbols are included in this backtest. Toggle click on the red and green arrows to turn on or off long and short bias filters. ie: if you have a green arrow showing then PV will only take long trades for that symbol. If you have a red AND green arrow showing, then PV will consider both long and short trades. You can specify a custom number of contracts to trade for each symbol separately using the + / - buttons. "a" means "automatic" and will use the gobal risk setting in "Basic Settings" section
CAPITAL is the amount of capital used in the backtest.
LEVERAGE There are two options of determining the # of contracts traded: "Equal contracts" or "Risk per bar".
"Equal Contracts" (new method) risk units are the number of contracts that are traded for each symbol. The number specified here will be the number of contracts traded for all 20 symbols. If you want to trade a different number of contracts for each symbol, then specify a custom risk unit in the "Set Symbols to Trade" area of the "Advanced Settings' section.
It is possible to trade 1/2 (.5) of a contract. When you set risk units to 1.5, then AlgoLab, and Performance Viewer will trade 1 contract 50% of the time, and 2 contracts 50% of the time (picked randomly) averaging 1.5 contracts over a period of time.
"Risk per Bar" (old method) It is the percentage (% / 100) of your capital risked per bar considering the time-frame being tested, symbol being traded, and recent volatility of the market. The risk value is used to determine how many contracts to trade based on the capital available.
PROFIT & LOSS The graph shows the total accumulated profit (sum) of all symbols, and it is this combined sum that is the basis of the system performance metrics.
Note that profit shown here is only the profit or loss resulting from your capital specified, and does NOT include that original capital amount. ie: the performance metrics do not include the effect of compounding returns.
SYMBOLS PL After your backtest has finished, you can view individual symbol profits. You can use the check boxes to quickly add or remove symbols and your backtest data will be updated immediately.
CUSTOM DATE RANGE After your backtest has finished, you can use the date range scroll bar at the bottom of the graph to quickly change the date range for your test. Your backtest data will be immediately updated
ANNUAL RETURNS TABLE This table shows each year starting capital (which, since we do not compound profits, is always the starting amount of capital), ending capital which is the amount of capital remaining after trading for that period, the profit which is the difference between the starting capital and ending capital (ie: how much profit or loss was earned during that period), the max equity which is the maximum equity during that period, minimum equity which is the minimum equity during that period, the max DD% which is the largest drawdown during that period shown as a percentage of the starting capital, and the annual rtn which is the total return for that year shown as a percentage.
MORE SETTINGS IN "ADVANCED SETTINGS"
DAYS TO TRADE Restrict trading to specific days of the week
COMMISSIONS AND EXCHANGE FEES are typical broker commissions and exchange fees. The default values are average Interactive Broker costs. You can input your own value.
SLIPPAGE is the number of ticks typically lost by stop orders and market orders being filled at a price x number of ticks worse than the last sale price listed in historical data. For back testing purposes, typically, slippage can amount to 1 tick per round trip - ie: 1 tick of slippage per buy then sell transaction. In the future, the # of slippage ticks will be adjustable.
REDUCER Since AlgoLab uses stop limit entry orders, there is a 10% to 20% chance that your entry order will not be filled due to a number of reasons including fast moving markets, or not enough buyers or sellers on the other side. The reducer amount refers to what percentage of trades to INCLUDE in the backtest, and trades are selected randomly each time you run or re-run a backtest. ie: 100 means that PV will always include 100% of the trades. 80 means that PV will exclude 20% of the trades chosen randomly.
MARGIN LIMIT Each position requires a minimum amount of capital in your brokerage account. If the total of the minimum margin amounts for each position are within this percentage of your total account value, then PV (and AlgoLab) will stop placing new trades until positions are closed.
CONTRACT QUANTITY CALCULATION
Equal contracts: This is the new default setting. This is a method of determining the number of contracts to trade per trade per symbol. It specifies an equal number of contracts per symbol. ie: if your risk is set to “2”, then you will trade 2 contracts per trade per symbol. If your risk is set to “1.5”, then you will simulate 1.5 contracts per trade with 50% of your trades as 2 contracts and the other 50% as 1 contract, thereby averaging 1.5 contracts per trade.
Risk per bar: is the old default setting. It is the percentage (% / 100) of your capital risked per bar considering the time-frame being tested. The risk value is used to determine how many contracts to trade based on the capital available. Higher risk value will trade more contracts which could result in higher profits, but also higher losses. Lower risk value could result in lower profits, but also smaller losses.
SYSTEM PERFORMANCE METRICS
P&L - this is the total profit or loss for the period
Gain/Pain - this is the maximum profit generated by the system divided by the maximum drawdown over the entire testing period. This is our favorite method of quantifying a system's performance, as the amount of profit that a system generates is less important than how much risk is assumed in order to generate that profit. Generally, the higher this Gain/Pain percentage, the BETTER the system performance is - that is, the higher this number is, the less risk is assumed per dollar of profit. Another way of looking at the Gain Pain ratio is it is the number of dollars earned in profit for every dollar lost in the largest drawdown.
Profit Factor - The Profit Factor is the ratio of the net profit versus the net loss. This ratio shows by how much the profit exceeded the loss. A value greater than 1 means the strategy has generated more profits than losses.
% Win - This is the percentage of winning trades. Note that the most profitable trading systems usually have a winning percentage of LESS THAN 50%. The strategies are profitable because the average winning trades are much greater than the average losing trades. A more important metric to consider is percentage of winning MONTHS or YEARS, not % winning trades.
Ave td - This is the average amount of profit or loss per trade.
# tds - Number of trades over the time period that this system generated
ave win - The average profit per winning trade
ave loss - The average loss per losing trade
win/loss - the average win divided by the average loss. If a system wins only 40% of the time (ie: 40% winning trades), but has an average winning trade that is 2 times the average losing trade, you can see how it can be an over all very profitable system.
max dd - A drawdown is the peak-to-trough decline during the period of the back test. Max DD is the maximum drawdown. This is an important metric to watch because it can give you an idea of how much money the system can lose over a period as great as the back test. In this example, the period is 9 years and the maximum drawdown was $14,026. On an account size of $100,000, that's a 14% decline which you could expect to happen at least once every 9 years.
max prof - The maximum amount of profit this system generated over the time period.
short profit - the total amount of profit or loss from only SHORT trades.
Long profit - the total amount of profit or loss from only LONG trades.
STERLING RATIO - The Sterling Ratio the average annual return is divided by the AVERAGE drawdown rather than the MAXIMUM drawdown. The Sterling ratio is an important metric to consider, as it quantifies the average return by how much drawdown can be expected.
SYSTEM ANNUAL PERFORMANCE METRICS
Ave rtn - Average annual return is the total profit over the period of the back test divided by the number of years in the back test shown as a percentage of the original invested capital amount. Note that AlgoLab return calculations are NOT COMPOUNDING. That is, the profit and loss is always withdrawn form the trading system, and not reinvested to increase the amount of invested capital. The amount of invested capital stays the same as the amount that you entered into the capital field all through the period of the back test. Compounding returns can have a positive effect on returns, but we chose to be conservative.
Ave DD - Average annual draw down percentage. This is the average of each years maximum drawdown shown as a percentage of the original starting capital.
Ave prof - Average annual profit shown in dollars
Max return - Maximum annual return.
Max an DD - Maximum annual draw down - This is the largest drawdown that occurred shown as a percentage of the original starting capital. This is an important metric because it will show you that you can expect at least one drawdown of this percentage once every number of years tested. In the example, we back test 10 years of historical data, so going forward, you should expect a drawdown at some point in the next 10 years of at least this percentage. Of course, your maximum drawdown could be more - or it could be less. This metric is also shown just below the profit and loss graph.
Max Prof - Maximum annual profit is the largest profitable year shown in dollars
Min rtn - Minimum annual return is the smallest annual return shown as a percentage.
Min DD - Minimum annual Draw down is the smallest annual draw down shown as a percentage of the starting capital.
Min prof - Minimum annual profit is the smallest annual profit shown in dollars.