This article suggests ways in which an investor might develop their own technical indicator.
Creating a new indicator is one of the most satisfying pursuits for those involved in the financial markets.
Not only can it teach you things about the markets which you would not have noticed previously but you might also discover something very valuable which could make you a lot of money!
Tools of the trade
Access to a computer will speed up the process. You will also need a charting package and a financial market data feed.
As far as software goes then Microsoft Excel is excellent for number crunching but if that is too expensive then you can also download free software from the likes of Opensource. A database package might also be helpful although it is not absolutely necessary.
A hand held calculator for quick calculations, a protractor for measuring angles and some rough paper will also be helpful. Also carry a small pocket note book to jot down ides in case inspiration hits you away from your desk.
You will need to be able to think originally about things and in this area I have found of great use the books of the writer Brian De Bono. De Bono is the master of lateral thinking, a technique for looking at problems from a new angle. He has constructed countless courses on how to ‘think outside the box’ and find new and creative solutions to old problems.
The value of observation
Your eyes are another important tool.
Close observation of market activity is an absolute necessity for anyone wishing to develop a new financial indicator.
Simply by looking at price action alone and running through screens and screens of charts can, I have found, yield amazing results.
For example, consider Welles Wilder’s RSI as shown in the chart above.
This indicator was based on Wilders’ observation that at certain points markets become overbought and at other times they become oversold.
He further noted that at these times markets tended to ‘rotate’ and begin moving in the opposite direction.
He developed the RSI indicator to measure oversold and overbought states in financial markets. It does this by calculating the relative difference between the strength of up-periods versus down-periods.
Another example comes from the analysts R.N Elliot, who discovered Elliot waves from carefully studying market behaviour over long periods of time.
Elliot was suffering from a serious illness when he made his discovery. He had taken to noting the hourly changes in the Dow Jones stock index out of a desire to kill boredom whilst he was recovering.
He drew up charts using the data and over a long period of time he began to notice recurring patterns in the charts which led him to formulate the Elliot wave theory.
I too developed a technical indicator from observation called the ‘Modified Stick Sandwich’ or MSS after a Japanese candlestick technique with a similar name.
The MSS was based on a recurring pattern I saw when markets were in an uptrend.
The pattern was made up of 3-bars in which the middle bar is a doji, or very doji-like, and it is sandwiched between two strong up-bars.
The chart above shows examples of the MSS in the uptrend and a downtrend and clearly shows how it is much more common occurrence in an uptrend and therefore a good indicator of bullish market conditions.
Just as Marcel Proust’s childhood memory of the taste of a madeleine cake became the starting point of his famous novel so our memories and experiences of the market can be the keys to unlock a whole rich store-cupboard of new insights and investment wisdom on which to base the invention of a new indicator.
Memories for example contributed to the development of one of my own favourite longer-term indicators – the AMR. It was inspired by the recollection of the speed and uncompromising rapidity of the sudden advance of the stock market in the spring 2009 after the sub-prime recession. I remembered how day after day the market kept rising despite continued widespread pessimism. This gave me the idea that it might be a principle of financial markets that they exhibit sudden initial bursts of directional movement at the start of new trends.
Another idea also came from my memories. It was the memory of making a lot of money and then losing a lot which gave me the idea of analysing equity curves using technical indicators.
I thought this way of approaching equity curves using the same analytical tools which analysts use to study charts might help to avoid a big drawdown in the future because it might be possible to identify a ‘top’ forming in the equity of an account and then take evasive action to avoid consequent losses.
Linking sunspots to soybeans may not be everyone’s idea of a worthwhile way to spend an analyst’s time but there are researchers who do just that in their quest to find the holy-grail indicator.
Whilst technical analysts tend to focus on raw price action there are other areas which can be researched for clues as to future market direction.
The sunspots mentioned above, for example, have a strong correlation with the price of commodities. Analysts have found that high incidence of sunspots, which are craters on the sun tend to be accompanied by higher commodity prices – even down to the price of vintage wines.
Back to Basics
Economic data can provide useful information about the future direction of market trends.
For example, we might stipulate that an increase in Money Supply could provide the basis of an indicator to forewarn of a change in stock prices.
In order to investigate whether this was true or not we might decide to compare changes in Money Supply with Stock Market data – but how could we actually go about doing this?
Data for the Money Supply of most countries is freely available on the internet and can be downloaded as a CSV file and then opened in Microsoft Excel. For the U.S the Federal Reserve website has money supply going back to 1959.
This can then be compared to stock market data which is also freely available on the internet in the same format.
The data can then be compared in Excel, and in this case in order to try to more clearly find a relationship the months in which money supply showed at least a 1.0% growth from one month to the next were filtered out into a list. This list was then applied to a chart illustrated below.
The chart shows the S&P 500 since 1955 with arrows pointing to the years which showed at least one month in which an increase of at least 1.0% in Money Supply was recoded. Above the arrow the number of months in which the Money Supply increased over 1.0% within that year was also recorded.
Looking at the diagram we can see that the years in which there tended to be a lot of money supply coincided with plateaus and troughs in the S&P 500. Note that between 1959 and 1967 there were no months in which money supply increased by more than 1.0% and likewise between 1986 and 1998. These were also periods in which the stock market rose at its fastest pace.
This seems logical when matched against our experience of economics as rise in Money Supply is often the by-product of more accommodative monetary policies adopted by nation’s central bank – and this is more likely in times of recession.
If we look closely we can see that during periods of high Money Supply there is never more than a single year – or sometimes very rarely a two year gap – in between years when the money-supply shows a more than 1.0% rise in at least one month.
This might lead us to formulate a rule based on our research that it is a good sign for the stock market if there are three whole years in which the money supply has not increased over 1.0% in any month.
This in turn could be made into an ‘indicator ‘for assessing good times to buy stocks.
So looking at one simple example which was worked up specifically for this article we can see that by using the wealth of free data available on the financial markets, important connections and correlations can be found which can from the basis of market indicators.
Developing a new indicator is often a slow process and it can take several attempts before success is reached. There is no point in creating an indicator which is not very useful so before publishing it test it rigorously.
As well as sequential back-testing try randomly back-testing the indicator it so as to make sure you are not just basing its success on a particular period in time, as secular market conditions can change and affect the usefulness of indicators.
Obviously this is sometimes not possible for longer-term indicators as a lack of historical data sometimes means only a set number of observable instances exist and this number is limited.
Also consider how a trader might implement the indicator as part of a trading strategy. The more you can ground it and make sure it is easy to implement the more successful it will be.
It’s no use for example saying that in certain situations there is a high probability the market reversing if it is difficult to trade. Try to put yourself in a trader’s shoes and make sure you can make money from your findings.
It is said that it took the inventor Thomas Edison 999 failed attempts to invent the electric light bulb and didn’t succeed until his 1000th attempt. Remember this anecdote when you are struggling to find discover the indicator which will give you an edge in the market.
You may try and fail many times before hitting gold.
Keep going and don’t give up hope – you never know one day you may be the next Welles Wilder or R.N Elliot, with your own book of indicators and hopefully a trading account full of money.