InStat Research developed out of a realization that quantifiable technical analysis is one of the best measures for evaluating price movement of financial instruments. The examples below describe a selection of completed strategy work.
M&M Strategy
Market breadthDaily scanSystematic strategy
This strategy trades in the direction of the breadth of the market, defined as the companies in the Russell 3000 index. Companies are scanned daily for movement greater than four percent and a moving average of the up/down ratio is created. When the moving average increases, the system identifies positive momentum.
The signal is used during upward-trending markets, measured by an occasionally optimized moving average. The strategy is designed to be cautiously traded and complementary to other strategies.
OR_RTM Strategy
Intraday futuresTrend followingMean reversion
This intraday strategy trades the 24-hour futures market and combines trend following with reversion to the mean. It is based on the observation that markets expand and contract on a regular basis, automatically selecting a reversion-to-the-mean approach after a large trend-following day.
When a market contracts, it may be ready to expand again. On expansion days, the strategy switches to trend-following mode, anticipates direction with an Opening Range Breakout system, and follows the trend until stopped by a parabolic stop or reversal.
ScareCrow Trading
End of daySignal generationError checking
This end-of-day system combines trending, momentum, short-term trend, and seasonal systems into two aggregate systems and then into one signal-generating system. Its outputs include strategy statistics, day-by-day signals, and error checking for data holes.
Automated Pivot Trading
Portfolio scanAutomated ordersVolatile markets
This strategy scans a portfolio of stocks every 30 seconds during volatile markets. When a price is close to a desired trading point, it places a multi-combination order to the broker, where the initial parent order waits as a limit order until its price is reached.
The workflow was designed to account for the latency that could make a completely reactive model too slow to place orders when internet, computer, and data lag are factors.