What is an Analytical Way to Go Through The Process of Selecting and Trading Investments
Executive Summary
- How does one select investments and manage trading?
- We will cover our method.
Introduction
Some of the most critical questions are how to select investments. In this article, we will cover a process we have developed.
*Note
This article is only part of our internal analysis of investing and communicates with people we know. We are not writing this to attract investors, and we are not giving investment advice. This is just a convenient way to document our analysis, which can be easily shared and easily updated.
Selection and Trading Step #1: How to Perform the Initial Selection
The first item or step is the initial selection. There can be many different ways to filter from the totality of the various options. However, whatever approach is used, there is a “basket” of potentials that seem like they will either go up (if one is going long in buying the actual item or buying an option for the item) or go down (if one is going short or buying an inverse ETF or using options).
A Brief Coverage of Investment Screening Software
Investment screening functionality, which is available in some trading software is efficient for this task. This allows one to add the columns that we think are predictive and then filter on the basis of these columns. Some of the software offers buy recommendations. One thing that is quite important is how long the buy recommendation has been made. Normally the recommendation will be for a Day, Week or Month. Normally, but not always, the longer the duration selected, the fewer the number of buy recommendations. We have decided to buy investment vehicles that have a longer duration, so a month. A reason for this is that we already have so many investment vehicles that we consider it a good thing to cut down on the number of investment vehicles we use.
Something we also found is that screening software does not offer balanced recommendations. That is, it tends to make more buy recommendations than sell recommendations. This is similar to investment banks and advisory firms. They normally have what are really sell recommendations camouflaged as “Neutral Recommendations.”
Selection and Trading Step #2: How to Determine Investment Timing
Much of the focus is often on the initial selection. However, the problem of timing is probably a more significant issue. One can easily see stocks like Tesla that are overpriced. However, timing speculating against overpriced stocks is very tricky.
How does one know when the bubble is about to pop? Investing means having insights and that other people agree with that insight. If you have an insight after everyone else has already had the insight, then you are too late, and you can’t profit from the insight. If you have the insight, but no one else agrees with the insight, or you are too early with the insight, you are wrong or too early and can’t profit from the insight.
The objective is to attain insights that others have not yet had and then make an investment decision that benefits from that insight. As others achieve the insights, the investment then pays off.
For this, one can use what is called investment timing methods. A very few investment timing methods are good enough such that while they are not right 100% percent of the time, they can be significantly more than 50% of the time. We have three different codings for any timing method.
- Effective
- Neutral
- Ineffective
We do not buy investment vehicles that our investment timing method does not rate as effective.
There are an enormous number of investment vehicles. However, no matter how much we like an investment vehicle, if we don’t have a demonstrated history of being able to forecast effectively, the timings of its price movements are up or down, we don’t buy that IV.
Investment Timing Option #1: Technical Indicators
Some well-known technical indicators are the following.
- Doji Star – Bearish
- Doji Star – Bullish
- Marubozu Black – White
- Marubozu Black – Bearish
Here is a definition of one of these technical indicators.
Dark Cloud Cover is a two candle bearish reversal candlestick pattern found in an uptrend. The pattern shows a possible shift in the momentum from the upside to the downside, indicating that a reversal might happen soon.
These technical indicators aim to produce a forecast, which translates into a buy or sell signal.
Technical Indicator Case Study Example #1: Test of Technical Indicators
We backtested around sixty of the most common technical indicators using around 20 ETFs and stocks.
We could find buy and sell signals that were sometimes correct. However, we could not find any of them that were good enough to rely upon. Because of this research, we don’t use technical indicators in our financial analysis. This is considered somewhat general knowledge among those experienced in trading. Normally technical indicators are combined with other types of analysis.
A Composite Technical Indicators
One can use multiple technical factors in a spreadsheet by setting up an AND statement to require that both values be a “go” before purchasing. And of course, we then review each indicator when making a sale.
Investment Timing Option #1: Financial Variables
Another type of variable is called a financial variable. This means that a value like price to book ratio or price to earnings ratio is used to perform forecasting.
The problem with using financial variables is that how the variable is valued changes depending upon other factors such as interest rates, group psychology, etc… Financial variables are very well reported upon, and financial variables are routinely quoted in discussions of investment vehicles without any commentary on the predictive ability of these financial variables.
How Market Fundamentalists Use Financial Variables to Produce Forecasts
If one could use financial variables to predict the price of investments with reasonable accuracy, the prediction process would be much easier than it is in reality. People that rely on financial variables are generally referred to as “fundamentalists.” They believe that the financial variables are what drive the valuation of investment vehicles.
We do not think this.
Using financial variables to produce forecasts sounds official and reasonable. We tend to accept those who use and can rattle off various financial variables without questioning the predictive ability of financial variables.
Financial variables are not sufficiently predictive to be used when it comes to predictive ability. And this is really the only important point.
Market Fundamentalists Versus Technical Analysts or Technical Traders
On the other side of the market, fundamentalists are technical analysts or technical traders. The market fundamentalist focuses on each investment vehicle and tries to know it inside and out. The technical trader does not care much about the investment vehicle per se. The investment vehicle could sell oranges or an ETF that goes up when a sector goes up. The technical trader cares about the line, not the vehicle. Usually, the technical trader may use technical indicators, but as scripting has become more common, mathematical scripts are just about to cover.
Therefore, both of these individuals are identified primarily through the tools that they rely on to formulate forecasts.
We will provide a case study of the forecast issues of relying on financial variables.
Financial Variable Case Study Example #1: Net Margin Versus Price Earnings Ratio
Let us take the example of Net Margin. This is a vital financial statistic or variable. However, if we review Tesla’s net margin, it is often less than 10%. Yet if we check Pfizer’s net margin simultaneously, it is often around 30%. This is historically high for companies to have this margin.
However, if we check the price earnings ratio between Telsa and Pfizer, Pfizer sells for around 20 times earnings, while Tesla currently sells for about 300 times earnings and has sold as high as 1300 times earnings.
The argument presented by Tesla proponents is that investors are buying “growth.” However, Tesla can never really grow beyond its high-priced niche, for several reasons we won’t get into here. And how could Tesla’s growth ever justify that multiple?
This is not to say that there is no relationship between financial variables and investment performance, and there is. It is just that the relationship is not strong enough to produce sufficiently accurate forecasts to use to choose or time the trading of investment vehicles.
Observation on the Case Study
For some stocks, investors are willing to pay one level for earnings, while others are willing to pay a completely different multiple. So what should be evident at this point is that we do not use financial indicators to forecast investment vehicles. However, we find financial indicators fascinating, and mostly in how investors value financial indicators so differently under different circumstances.
Investment Timing Option #3: Mathematic Scripts
Mathematical scripts use several inputs to drive prediction. Anyone can create a script and load it into software, and they are so high in their variance that it is difficult to provide a specific definition of what they are. Mathematical scripts have a significant advantage over technical indicators in that they allow for much more flexibility. Many mathematical scripts have parameters and therefore function as a statistical forecast model. This means the parameters can be adjusted to test and improve forest accuracy under various circumstances.
Univariate Versus Multivariate or Regression Mathematical Scripts
Most technical mathematical scripts are univariate predictions, and the mathematical scripts just use the history (like sales history in product forecasting) to produce the forecast. If one can find a correlated variable, one can produce a multivariate or regression or even an AI type indicator. One of the most commonly used variables is interest rates, and as interest rates decline, assets tend to appreciate.
Surprisingly, the lack of applicability or usefulness of multivariate forecasting applies to product forecasts. There are continual promotions of using multivariate forecast models in product forecasting, with AI requiring many variables.
The problem is that while variables can be found that have relationships with the item being forecasted, many of these relations turn out to be spurious.
The Ceaseless False Promises of Multivariate Forecasting
The promises of multivariate forecasting methods have been unfulfilled for decades. When AI became very popular around the mid-2010s, a fresh crop of predictions was made for the significant enhancements that AI and multivariate forecasting would bring, which again turned out to be false.
Mathematical Scripts Case Study Example #1: Test of Mathematical Scripts
Copying our testing of technical indicators, we backtested around sixty of the most common mathematical scripts using around 20 ETFs and stocks. The vast majority of them we did not find helped us make forecasts. This is particularly true because we look for assets that have appreciated too much and are due for a correction. However, bubbles can continue for long periods.
A good example of a mathematical script is the Volume Weighed MACD Histogram. This is covered in the article Evaluation Of A MACD Investment Forecast Script.
Selection and Trading Step #3: Determining The Optimal Number of Investments to Hold
Charles Munger of Berkshire Hathaway has stated that one is lucky to have four suitable investments and critiques diversification. The general reason for diversification is it helps mitigate risk.
When it comes to the topic of the investment vehicle number, the reason for having several investments in different vehicles is how often our investment vehicle selection method is accurate. All investment vehicle selection methods that we are aware of must have an error level. That is, the investment vehicle selection method says the investment vehicle will go up, and it goes down or stays flat, and the converse.
An investment vehicle selection method can be found that is more accurate than not. But none of them is 100% or close to that. However, the more investment vehicles owned, the higher probability the investment vehicle selection method will be correct. On the other hand, the more IV one buys, the more difficult it is to keep track of them.
Using the proper mathematical script means having enough investment vehicles to benefit from the predictive power of the script. This is similar to the logic of how casinos try to make people stay in the casino. A person can beat the odds in the short term (or in limited playing time). However, the casino knows the longer the person plays (and the more the number of bets accumulates), the more the player’s return will match the average odds of the game, which are stacked in the casino’s favor. This means that it seems that just a few investment vehicles will not allow one to take advantage of the predictive power of the investment vehicle selection method.
The side that opposes this is that if one has so many investment vehicles, it becomes difficult to keep track of them, and it is very important to track the investment vehicles. We would prefer to not have more than 20 ongoing investment vehicles. If one is planning on buying and holding, then the number of investment vehicles can be higher, as the maintenance is lower, but we do not follow a buy and hold strategy.
Conclusion
All of this is part of creating a repeatable and controllable process for a disciplined way of selecting and trading investment vehicles.