Here is an explanation of the BiMTi number that I’ve been posting on my 60-Second Updates.
History: A few months ago, when I was contemplating leaving Westrock Asset Management, I realized I needed to reshape the content I had been writing for internal use to accommodate for a broader audience, one that may not have been as familiar with the factors that at times we discussed in the office.
I also was focused on creating something that could very concisely tell my view on these markets. My “view” or my “opinion” on the markets is not a trading signal. My view is simply a lightly painted canvas that colors the background of my trading decision. When different factors line up properly, I know that I need to press on the gas and take some risk. Otherwise, as it is most of the time, things are just floating in a range, somewhere in the middle, without a precise call to action.
I had and still don’t have any idea if this is a reliable tool for other traders. I know that for me, it just helps me to summarize my thoughts and give me a little green, orange or red light before pulling the trigger on a trade. I picked a 6-10 week range, because I think that the “backdrop” canvas typically affects prices over that period of time. Occasionally there are new pieces of information that change the balance dramatically (say, a hurricane or policy surprise decision), but those situations are rare.
On of the best byproducts of the effort to create this weekly indicator is that it gives me the discipline to go through all the information available to evaluate trade ideas and prepare for the week ahead. Similarly to the way my blotter works, the process of creating the indicator helps me pause, block the intraday noise, and focus on the bigger picture.
Ingredients: Here is what I put in.
- Supply and demand situation
- Weekly chart
- Momentum indicator (RSI – 9-day)
- Commitment of Traders
- “Gut” number
I give each component a number between +5 and -5. This process is pretty subjective, but I try to be consistent in my subjectivity. Meaning…
…I will give a large pattern breakout a 4 or 5 on the weekly chart input every time I see one. While I will give a 0 if it’s in the middle of a range. A -4 or -5 for a breakdown of a large consolidation.
…A tight supply and demand situation calls for a high positive number, while an oversupplied market, calls for a negative number. I don’t forecast the S&D. I stick to what information is currently available on the present situation, which typically impacts the market for at least 2-3 months.
… I distribute RSI numbers normally. That is, RSI of 100 = +5; 50 = 0 and 0 = -5. (I’m still working on this one. I’m not sure I’m manipulating this one well…)
… I look at non-commercial positions in the legacy Commitment of Traders. I consider the numbers in relation to their record long and record short positions. A net position (i.e., longs minus shorts) that is near record level would be a negative number if near record long and positive number if near record short. This is a mean-reverting input and tends to counter-weigh the RSI data. It is also a weekly, delayed input (data reported on Friday is recorded the previous Tuesday — by the time I consider it on Sunday afternoon, it’s still valuable from a big picture, historic perspective, but not for immediate impact) so I use it carefully.
…”Gut”. Yeah, that gut thing is just something that I’ve learned to trust. It’s tough to give it a number, but after reading as much as I do and looking at charts for thousands of hours, I’ve developed a filter that I need to respect. It’s not consistent. And it had led me to lose money at times (more, opportunities to make money). But I typically sleep better when I listen to it.
I weight each of them uniquely by a multiple of 10. I put more emphasis on momentum as opposed to the COT, namely multiplying the former by 30 and the latter by 10. The weighted average is the the BiMTi. Simple. Not rocket science. Not exactly art, either.
Future: I’m waiting for several months of data to backtest and see how accurate it was in predicting the general price direction (again, NOT a trade signal. Signals need to go with a specific risk management plan, not offered here). I suspect I’ll need to improve the timeframes I use for the indicators, because they aren’t consistent. S&D has a long time horizon, while the RSI is much shorter. I’ll probably tweak them over time. For now, here you have it, my own proprietary indicator to help us make quick judgements about the medium term outlook of a market.