SPY's Quarterly Rebalancing Pattern?

Archives
1

This is a look at the idea that today’s sell-off on zero news was caused by a quarterly rebalancing of funds. What I mean by this is that if a fund is supposed to maintain 80% stocks and 20% bonds for example, and stocks outperformed that quarter, then they would rebalance by selling stocks and buying bonds to restore the 80/20 balance.

Here I’m looking at the last two trading days of the past quarters, and the following first day of the next quarter. I am using 15m candles and then 30m candles beyond Nov 1 2021 because TradingView won’t let me see past that point with 15m candles :frowning:

Q1 2022 | 2022/03/30 - 2022/03/31
9.50 move down: 461 → 451.50

image
image1022×372 6.23 KB

Q4 2021 | 2021/12/30 - 2022/01/04
4.30 move down: 479 → 474.70
And then gap up, sell off, and rally

image
image1123×343 6.76 KB

Q3 2021 | 2021/09/29 - 2021/10/01
7.90 move down: 436.90 → 429.00
And then open green, sell off, and rally

image
image919×250 3.54 KB

Q2 2021 | 2021/06/29 - 2021/07/01
0.25 move up: 427.90 → 428.15
And then open green and rally

image
image1283×348 6.23 KB

Q1 2021 | 2021/03/30 - 2021/04/01
2.25 move up: 394.00 → 396.25
And then gap up and rally

image
image1257×348 5.84 KB

[size=4]Commentary:[/size]

  • Q3 and Q4 2021 had the same pattern of two red days followed by a green. So far, Q1 2022 is looking similar, and we will see how it goes tomorrow to confirm if it will be a green day.
  • Q2 2021 was the most stable with less volatility and a strong first day to the next quarter.
  • Q1 2021 was interesting with a flat day, volatile day, and then a rally.
  • Consistency?
    It seems to be that generally (but a bit of a reach) on the last two days of a quarter from about 10 AM PST ish to 11 AM PST ish, volatility begins and it looks like larger sell-offs occur especially closer to EOD (except at the end of Q2 2021).
  • A clear consistency that we can spot is that every time, the first day of the new quarter rallied upwards intraday.

[size=4]Questions/Notes:[/size]

  • Each quarterly transition was dealing with its own set of macroeconomic news. Are the patterns just coincidence?
  • Do funds even rebalance quarterly? Maybe they do it every day? Or every week? Perhaps each do it differently. If so, are there enough funds that rebalance quarterly to create patterns like this?

[size=4]Trading Takeaway:[/size]

  • The main consistency here is that on the first day of the new quarter, it could be a volatile opening, but generally the intraday will be bullish. We’ll see.
5 Likes
2

Idk if I’m getting the quarterly dates right, but typically the day after the quarter, 4/1, 7/1, 10/1, 1/1, these days have all ended green the last 4 times. Assuming this trend holds, tomorrow should be green with volatility at open. But the trend after is dependent on the market conditions. Just seems like the day after is green but after that is a wild card

2 Likes
3

Those are the correct dates.

And that is what I got out of this analysis as well.

2 Likes
4

Certainty is an interesting theory and definitely some truth to it. I saw bonds post a nice rally on seemingly no news as well these last couple of days it could make sense. I’ve grabbed some treasury puts again after what looked like a pretty classic pivot chart on the 10 years. I’ll be watching spy today as well and see if we see some green. I also witnessed the bond rally start live it was fast, there was also huge selling right before it dropped like 0.20% in seconds then recovered out of no where

2 Likes
5

I thought bonds go down when they get bought?

Taking a quick snapshot, we see that the 10-year has been going down this week.

image
image1058×630 53.4 KB

1 Like
6

Great stuff, Kevin!

It’s fascinating that this is a pattern that repeats itself, as presumably brokers are supposed to their best to diffuse impact of such large transactions, especially by using the dark pools and doing so over days. Perhaps the magnitude of rebalancing is so large that it is impossible to hide.

Your finds are corroborated by the FT, and there’s some additional context and historical evidence to boot!

By the way, the size of the equity market is ~$50T while bonds are about ~$120T. Bonds are also the smaller portion of these funds. Perhaps that is one of the reasons the effect on bonds is not seen as much - it’s diluted over a much larger pool than equities.

3 Likes