Summary
For last week I’ve been neutral to slightly bearish as written (link) seven days ago. Until Tuesday the market showed a considerable pull-back (2.15%) . Though the week showed a gain of 0.19%.
Setup for week 11 is bullish, hence I expect a higher close for the week. Furthermore I’m expecting a mild pull-back in the range of ~ 1%.
Price Action
[bullish]
- Direction: long term up, intermediate-term up
- Volatility: low, volatility expanding
- Trend-health: long-term trending environment (TSI >1.65) and intermediate trend conditions have weakened for twice over the last two weeks.
- Channel: intermediate channel position very low.
- OB/OS: RSI(5)>50 and <75, RSI(10)>50 and <75
.
Seasonality
[bullish]
Seasonality is bullish for March (K) as well as for week 11 of the year (L).
.
Correlation
[bearish]
Correlation (M) among S&P500 members remains low (<0.25) . Short-term we saw a significant expansion.
Read more about my correlation related research here.
.
Breadth
[bullish]
Market Breadth (N) for top performers is at high absolute levels and decreased over the week, indicator value is at highest quartile. In the past this has been a highly positive setup.
.
Sectors
[bearish]
Sector setup (O) is bearish. The sector leaders are XLI (Industrials) and XLF (Financials).
.
Bonds
[positive]
RISK ON for the bond market (P). Generally this is positive for the stock market, especially with a combined setup of DIRECTION & VOLATILITY.
.
I find these weekly updates quite helpful. Thank you, Frank.
Hello John,
thanks for your feedback. Let me know if you have an idea how to further improve this.
Frank
Frank,
Great work. A few questions: have you looked at lead lag relationships? have you looked at normalising the inputs based on some “importance weight” index? Some of these factors I have doing work myself so I follow them explicitly. happy to discuss offline. Also, I would be interested on your thoughts on use of volume as well as “risk aversion metrics” (put call ratios, Bull/Bear Indicators etc.). Great work, great site, keep it going!
Hi Frank,
Are you able to explain the diagram in your summary session?
Cheers,
Ellis
Hello Ellis,
each test (setup) is having a unique identifier ( A – P ). The setups undertaken are described in the text underneath the diagram. As price action has multiple dimensions (direction, volatility, trend-health, overbought/oversold, volatility) i had to add a second dimension to the picture. When you see multiple X per column than this simply means it’s a combined setup, .e.g. direction AND volatility.
Underneath of the setup area you find the associated statistics for each of the setups.
Statistics: each week represents one trade, from Monday open to the week after Monday open. Win%: percentage of trades closed higher than it opened. Avg%: average percentage return. Sharpe ratio of trades is defined as: annualized average percentage return of trades divided by annualized standard deviation of returns (trades).
Regards,
Frank
Hi Frank,
If each week just represents one trade, why the statistics changes quite a lots every week?
You comment is useful to understand the market but I can’t relate it to the diagram at all.
Regarding the diagram,
- are you using the setup in your portfolio?
- which setup used in which portfolio?
- how to allocate, and which setup is active?
- why they used in that way?
Cheers,
Ellis
Hello Ellis,
each weekend i take new setups ..columns …..depending on state of the market. With my setups I try to “describe” past market action and thereby look at the outlook for next week.
E.g. Rate of change 252 bars > 0 and intermediate trend up. This happened about 465 times in the past (column B). The expectation for the week after is 0.18% -> this is just one example.
I then do that for a number of different factors,e.g. leading sectors, volatility, bond market, correlation etc. This gives me an idea of the technical picture going into the new week.
It’s kind of “Situational Quantitative Analysis”.
So this is not a trading system, but it helps to asses how to play the coming week.
Regards,
Frank