|Asset Management, Consulting, Financial Model, Financial Services, Personal Finance, Stock Market|
|Excel, Financial Analysis, Financial Model, Financial Modeling, Investors, Performance Tracking, Profitability Analysis|
Designed to support medium-term investors who hold onto positions for days, weeks, or even months. Not intended for day traders or scalp traders. Best suited for investors who do not wish to be tied up to a computer screen for hours but limit their involvement to a few minutes each day, trading closing prices.
Looking back, since 2008 the model has shown remarkable performances in terms of annualized profit for some particular stocks in different markets. To name a few: Siemens 33%; Intesa Sanpaolo 30%; JP Morgan 25%; DTE Energy 25%; Coca Cola 20%; Conoco Phillips 20%; Marubeni 18%; Toyota 16%.
The model uses a well-balanced mix of 6 among the most popular technical indicators, equally divided among Volatility, Trend, and Momentum categories. They are:
- Bollinger Bands
- Simple Moving Average
- Pivot Points
Each indicator generates specific trading signals when prices reach pre-determined levels. A final BUY or SELL signal is created according to parameters that vary depending on whether the selected stock is cyclical or non-cyclical, to take into account different price dynamics. Cyclical stocks: when at least 3 indicators (1 per each category) get activated. Non-cyclical: when at least 2 indicators (1 per category in any 2 out of 3) are on. Stock cyclicality is assumed by default based on average historical price volatility, but it can be changed if desired.
The model uses historical daily prices imported by the user (going back 15 years max; should not be less than 10 years) plus user-specified technical indicators levels, to produce a simulation of historical trading returns (backtesting) based on the assumption that BUY and SELL occur on the first such signal. Stop losses are factored in based on the original buying/selling price assumed by the model. If short selling (shorting) is indicated as an option, positions opened on both ends, short and long. Otherwise only long.
Such historical trading results are compared to stock appreciation over the same term, to evaluate the profitability of trading strategy vs. buy and hold strategy.
If stock performance exceeds trading performance in most years, BUY & HOLD may be a better strategy than trading, for that particular stock
The trading strategy built into the model is simple, designed to produce the best results over the medium term with limited time involvement: SELL when the price breaks above-defined range and BUY when it breaks below. To decrease the probability of false signals leading to untimely sales or purchases, range levels can be modulated through “buffers” (different for BUY and SELL) modifying the ability to produce signals.
With particular stocks, sometimes better results can be achieved by using a “blank”, to neutralize the impact of a particular indicator or buffer. Other times, changing a buffer may not produce any change in the result if, for example, the signal is duplicated by another indicator in the same class.
Returns are shown in both, trailing and calendar year fashion. Other information provided includes: average holding time of position, idle time between positions, realized and unrealized gains, winning trades %, win/loss ratio, profit/loss ratio, average profit per trade, etc.
Once all technical indicators have been tuned up to generate the best possible historical result for the desired lookback period, the idea is to keep on updating prices daily, executing trades according to the model’s instructions. Besides target price, the model indicates if we are opening or closing a position and if the closing is due to stop-loss kick-in.
Even though price updates can be done (and signals obtained) at any time during the day entering current price as of closing price, to maintain the model’s integrity (trading signals are based on historical closing prices and volumes) and fully exploit its potential, it is best to trade closing price, choosing either one of the following options:
- 15min before the end of the session, enter the current price as the closing price in the ‘Data Import’ sheet and operate based on the signal.
Normally, at this point, the price should not be far from the actual closing price to be determined during the closing auction.
- Wait for session closing, enter the closing price, and trade, based on the signal, during the after-hours session or at the opening of the next day’s session.
Obviously, the closer your actual trading price will be to the closing (target) price, the more reliable the information obtained through the model. For best results over time you should always act according to instructions, trying to avoid second-guessing the model in the attempt to maximize profits or reduce losses. When you first start using the model, if the first signal you get is a BUY or SELL signal for closing a position opened earlier, you should not operate and wait for a new cycle to start with an open position signal.
OPERATING STEPS RECAP
- Input trade amount, trading style, and time horizon
- Upload historical prices (as described in separate instruction document)
- Select program and technical indicator parameters
- Analyze historical returns
- Fine-tune parameters (as described in separate instruction document)
- Update price daily and operate based on a trading signal
TO KEEP IN MIND
No investment strategy generates profits all the time under all market conditions. Some years you earn and some years you may lose. Since this model is designed to produce the best results in the medium term, what really matters is the global result over your entire investment horizon (should be at least 3-5 years).
Risk cannot be eliminated, but only mitigated through a solid and consistently applied trading strategy over an adequate time horizon. Therefore, second-guessing the model’s signals trying to maximize profits (or avoid realizing losses) is not advised, since it might not produce the best results in the end.
While technical indicators look at price history on the assumption that similar price dynamics will repeat themselves in the future, there is no guarantee that past performances may be repeated.