One process you can use that doesn't depend on any specific history of events is to use simulation and if-then analysis instead. Simulation allows you to evaluate all possibilities.
Some examples how you can use simulation and if-then analysis outside and in combination with your discretionary or system trading specific analysis:
* Look at biggest up/down day or periods over given period to determine risk
* Look at biggest up/down days over a given volatility level
* Look at the risk the perfect trader might take.
* Look for longest stretch of up/down days in market/strategy. Autocorrelation could be a significant risk.
* Create synthetic data series using different projections of volatility to see how your strategy does
* Try to find predictors of trend vs range markets. Use a combination of those predictors and price action to trade markets in ways that dont rely on historical analysis.
* Chop up the data
* Define your trade/prediction as a series of many forms and diversify over each form to account for every possibility that you consider a win. This solves the "bounding problem". It also solves the information less statement problem.
On the last point, let's imagine you are very good at predicting the market in the short term from 1 to 3 days. The problem is that if you convert his into a trade, you are likely to introduce arbitrariness. The arbitrariness can derail your plans. Some forms of arbitrariness are difficult to solve for or simply ignored. As an example, if you open an account a broker that happens to be engaged in some form of fraud. Simply the broker you chose becomes an important predictor of your success. It shouldn't be. Or say you think the market will rally and you buy an ETF but something goes wrong with the pricing mechanism and the ETF doesn't track to the market. Again, randomness stymied you.
However, some types of arbitrariness can be better accounted for. For example, let's say you predict the market to rally on day 1. You could buy the futures or options. Futures exposes you to path risk. Options endpoint risk. Dividing up your trade into multiple part executions is "diversification over forms". It allows you to capture the majority of "success forms".
Example, imagine you imagine the market to rally... you could place a large trade with a tight stop. The return could be high if your right. But, if the tight stop is arbitrary you've introduced randomness. Dividing up the capital for the idea into several "trade executions" should diversify. As an example, let's say you choose to use a large stop but the market really tanks the first day but on the second rallies above the first. You could open the first day with a large stop but say leave 30% of the money for the position to re-enter on the second day.