A couple of weeks ago I promised the board I was going to do some analysis on Quad Witching (QW) which if you're not familiar is the third week of every third month, so the third week of March, June, September and December. This is the week when we see stock index futures, stock index options, stock options and stock futures all expire at the same time. I think it's a commonly held belief that we see a lot of price manipulation during monthly option expiration weeks and that QW weeks is that manipulation on steroids. The purpose here was see if that perception is correct and if it is to find a tradable advantage.
For my part I believe prices on key stocks such as AAPL are heavily manipulated around Opex weeks and that QW represents a massive effort to transfer cash from Joe Retail Trader to the institutional trading companies. With that in mind I fully expected to find that regardless of market direction around QW that I'd see heavy negative price action during the week of QW. I really didn't set out to do a chart analysis but rather a statistical one, meaning I went in expecting to see QW week show me a disproportionate number of down weeks v up week when compared to the other weeks in the analyzed timeframe. That is not what I found.
Approach:
I downloaded 10 years of weekly S&P500 data from Yahoo Finance. It is worth noting that the entire period of time has been bullish so if QW saw a higher portion of down weeks than average it would stand out fairly clearly. I manually labeled each third Friday as E, each third third Friday as QW, the week prior and the week after QW as P and A. I THINK the P and A weeks will lead to another post, hence the Pt1 in this thread title. This gave me 40 weeks of QW to analyze and I did most of that using 100% column charts, meaning a positive week was up, negative down. No real effort to analyze the size of the movement. That may come later but it's a bit of a different effort.
Results:
Well, long story short I was wrong in my basic hypothesis but that doesn't mean you should stop reading here. There is most definitely a tradable advantage here it just took some time mentally reconciling what I believed I would find with what my data was showing me. The basics results were for the 523 weeks I analyzed were:
The above percentages were consistent for all 10 years and the first five and last five were similar. The only hint of price manipulation appears in the last three years when we see a literal coin toss for whether QW week is up whereas the overall market for the same three years is up 61.1% of the time. That's nice info but IMO not tradable. I want better than 6:5 odds if I'm going to act on this data.
This is where I found myself stumped as I realized that for the 10-year period QW actually sees a slightly higher percent of weeks increase that the overall market. That was completely counterintuitive for me. After setting it aside for a few hours I began to wonder about Friday price action and this is where this gets interesting. So I downloaded the same 10 years of daily S&P 500 data from Yahoo Finance and the picture and the tradable advantage began to appear.
So to not bury the lede Fridays are HEAVILY manipulated during week of QW. Over the past 10 years Friday is actually the most bullish day of the week with 57.1% of Fridays closing higher than the prior day. The next closest day is Wednesday with 55.7% of Wednesdays closing higher than the prior day. So what do we see on Fridays of QW:
That last three years is really 2/13 and all of the prior six and 8 of the past 9 QW Fridays have been down. Yep, you've read that right. Despite Friday being the overall most bullish day of the week for the past 10 years over the past three years only two QW Fridays are bullish. Chart below shows QW weeks for past 10 years, again just up or down. You can "see" Friday getting more and more manipulated as time passes.
It is also worth noting that over the past three years the bullish nature of each day of the week has changed. A quick comparison of 10 years v 3 years by day. Percentages represent days that were Up.
M - 53.5% - 57.2%
T - 54.0% - 51.0%
W - 55.7% - 58.2%
R - 55.3% - 59.5%
F - 57.1% - 56.3%
So Friday is slightly less bullish and has moved from most bullish to fourth, Monday, Wednesday and Thursday have all become considerably more bullish and Tuesday has become a coin toss. None of that is really tradable, IMO but rather just interesting info and good context for the fact that over the same 3 year window where we see 56.3% of Fridays end up we see only 16.7% of QW Fridays end up.
OK, last piece of data on Fridays. If you're still here you're a sick mind. Like me! This has actually been evolving as I write this post. QW Fridays over the past three years look like a bull trap. Combine the fact that over the past three years Wednesday and Thursday are the most bullish days of the week with the fact that QW Fridays gap up 62.5% (7/12) of the time but trade up only 8.25% (1/12) times and end down 83% of the time. Run up into Friday then crash on Friday. The last QW Friday that ended up was March 15, 2019 so every QW Friday for over two years has been red. Wow.
Conclusions:
Prices are definitely manipulated the week of QW to set up a bull trap on Friday. The rest of the week looks about like non-QW weeks. Fridays however are incredibly heavily manipulated and it appears to be increasing over the past three years to the point where it's become tradable. When less than 10% of QW Friday candles are green to me that's a real advantage!
OK. That was a mouthful. Eager to hear the board's thoughts.
For my part I believe prices on key stocks such as AAPL are heavily manipulated around Opex weeks and that QW represents a massive effort to transfer cash from Joe Retail Trader to the institutional trading companies. With that in mind I fully expected to find that regardless of market direction around QW that I'd see heavy negative price action during the week of QW. I really didn't set out to do a chart analysis but rather a statistical one, meaning I went in expecting to see QW week show me a disproportionate number of down weeks v up week when compared to the other weeks in the analyzed timeframe. That is not what I found.
Approach:
I downloaded 10 years of weekly S&P500 data from Yahoo Finance. It is worth noting that the entire period of time has been bullish so if QW saw a higher portion of down weeks than average it would stand out fairly clearly. I manually labeled each third Friday as E, each third third Friday as QW, the week prior and the week after QW as P and A. I THINK the P and A weeks will lead to another post, hence the Pt1 in this thread title. This gave me 40 weeks of QW to analyze and I did most of that using 100% column charts, meaning a positive week was up, negative down. No real effort to analyze the size of the movement. That may come later but it's a bit of a different effort.
Results:
Well, long story short I was wrong in my basic hypothesis but that doesn't mean you should stop reading here. There is most definitely a tradable advantage here it just took some time mentally reconciling what I believed I would find with what my data was showing me. The basics results were for the 523 weeks I analyzed were:
- 60.2% (315/523) of all weeks were up
- 47.5% (19/40) of Prior weeks were up
- 40.0% (16/40) of After weeks were up. This and the above bullet are interesting
- 51.3% (20/39) of QW weeks gapped up Monday morning
- 60.0% of QW weeks traded up. This was the biggest surprise. I expected a G-R week and that's not what I got.
- 62.5% of QW weeks were up.
The above percentages were consistent for all 10 years and the first five and last five were similar. The only hint of price manipulation appears in the last three years when we see a literal coin toss for whether QW week is up whereas the overall market for the same three years is up 61.1% of the time. That's nice info but IMO not tradable. I want better than 6:5 odds if I'm going to act on this data.
This is where I found myself stumped as I realized that for the 10-year period QW actually sees a slightly higher percent of weeks increase that the overall market. That was completely counterintuitive for me. After setting it aside for a few hours I began to wonder about Friday price action and this is where this gets interesting. So I downloaded the same 10 years of daily S&P 500 data from Yahoo Finance and the picture and the tradable advantage began to appear.
So to not bury the lede Fridays are HEAVILY manipulated during week of QW. Over the past 10 years Friday is actually the most bullish day of the week with 57.1% of Fridays closing higher than the prior day. The next closest day is Wednesday with 55.7% of Wednesdays closing higher than the prior day. So what do we see on Fridays of QW:
- The past 10 years we see 40% (16/40) of Fridays close higher
- Over the past 5 years we see 30% (6/20) of Fridays close higher
- Over the past 3 years we see 16.7% (2/12) of Fridays close higher
- Over the past year we see 0% of Fridays close higher
That last three years is really 2/13 and all of the prior six and 8 of the past 9 QW Fridays have been down. Yep, you've read that right. Despite Friday being the overall most bullish day of the week for the past 10 years over the past three years only two QW Fridays are bullish. Chart below shows QW weeks for past 10 years, again just up or down. You can "see" Friday getting more and more manipulated as time passes.
It is also worth noting that over the past three years the bullish nature of each day of the week has changed. A quick comparison of 10 years v 3 years by day. Percentages represent days that were Up.
M - 53.5% - 57.2%
T - 54.0% - 51.0%
W - 55.7% - 58.2%
R - 55.3% - 59.5%
F - 57.1% - 56.3%
So Friday is slightly less bullish and has moved from most bullish to fourth, Monday, Wednesday and Thursday have all become considerably more bullish and Tuesday has become a coin toss. None of that is really tradable, IMO but rather just interesting info and good context for the fact that over the same 3 year window where we see 56.3% of Fridays end up we see only 16.7% of QW Fridays end up.
OK, last piece of data on Fridays. If you're still here you're a sick mind. Like me! This has actually been evolving as I write this post. QW Fridays over the past three years look like a bull trap. Combine the fact that over the past three years Wednesday and Thursday are the most bullish days of the week with the fact that QW Fridays gap up 62.5% (7/12) of the time but trade up only 8.25% (1/12) times and end down 83% of the time. Run up into Friday then crash on Friday. The last QW Friday that ended up was March 15, 2019 so every QW Friday for over two years has been red. Wow.
Conclusions:
Prices are definitely manipulated the week of QW to set up a bull trap on Friday. The rest of the week looks about like non-QW weeks. Fridays however are incredibly heavily manipulated and it appears to be increasing over the past three years to the point where it's become tradable. When less than 10% of QW Friday candles are green to me that's a real advantage!
OK. That was a mouthful. Eager to hear the board's thoughts.
A fearful society is a compliant society. That's why Democrats and criminals prefer their victims to be unarmed. Gun Control is not about guns, it's about control.