Key Trading Ideas: Trade Volatility, Trade a Seasonal A Simple Crude Oil System, End-of-Day: Illustrates the ideas of trading volatility, trading a seasonal and trading yesterday Buy Monday, close < yesterday’s close range of last 4 closes is less than the average true range for the last 4 days Sell Monday, close > Read more about [idea] Third Crude Oil System #idea #crudeoil #101tradingideas[…]
Key Trading Ideas: Trade relationships, Trade a fundamental, Trade Yesterday, Trade a Seasonal A Simple Crude Oil System, End-of-Day: With the ideas that we are two days after the inventory announcement, Corn is an oil substitute, and Ten Year Notes reflect risk to the economy. Buy Friday, close > yesterday’s close close in Corn > yesterday’s Read more about [idea] Second Crude Oil System #idea #crudeoil #101tradingideas[…]
A Simple Crude Oil System, End-of-Day: Buy when close > close 9 days ago + 5 std dev Sell when close < close 9 days ago – 5 std dev Eligible to reverse after 2 days in trade* Exit after 20 days in trade Results, 2006-p: 19 of 34 wins, 55% Avg trade: $1.88 Read more about [idea] Crude Oil System #idea #crudeoil #101tradingideas[…]
HFT Algorithm Modern video games are designed, or at least vetted, by psychologists. The idea is to make the game challenging enough but not to challenging so players give up. Use the same principle to design an HFT algorithm: Trade where the capitulation points are, Trade where participants are likely to ‘continue playing’ Trade the liklihood of Read more about [idea] HFT Pain Algorithm #idea[…]
Forecasting Crude Oil Looking back over the week, the intraday correlation of Crude Oil and ES is very high. Which means that forecasting one will let you forecast the other. A cool use of high correlations.
Trend Days and Regime shifts are similar phenomena. See Standing Ovation Model and Granovetter’s Model. For the last 90 days I’ve been forecasting Trend Day or Rangebound Day on the prior day’s close from historical data. I get the direction right about 62% of the time, but Trend Day/Rangebound Day results only about 50% correct. Coin flip odds. Read more about [Idea] Trend Days and Regime Shifts are Similar[…]
Continuous. Discrete. Linear. Cyclical. New Geometries = New Insights Plot a market using a different geometry. I was trying to plot on a circle but ended up just using the same plot and different colors for each day of the week. The last few days look a little rangebound. Read more about [Idea] New Geometries, New Insights #es #spy[…]
Questions that lead to ideas…
Market Models: Discrete, Continuous or both? Coursera has a great course in models: Model Thinking
Trading Ideas: Automated Japanese market model Sounds a lot like the winning model in a Kaggle competition. Interesting overview of components and methodology.