Stock Price Variance
through January 20, 2022

One way to measure the riskiness of US-listed stocks is to measure the variance of stock price changes. Risk is a function of stock price variance, which is the degree which stock prices vary around their mean change. We believe that risky stocks with high variance require investors to demand a higher expected return to hold them. The high uncertainty presented by higher variance requires a higher expected return.

As of January 17, 2023, we have reached a plateau of stock price variance. It is elevated, but no longer rising.

We calculate variance based on daily, adjusted price changes over the past 253 trading days. Stocks that maintain a stable price or stocks that rise (or fall) steadily have lower variance than stocks that fluctuate aggressively.

This is counter-intuitive in practice. As stock price variance rises, we expect to see stock prices fall, which increases expected returns for new investors.

There are many experimental questions and hypotheses around this topic:
1. What causes increased stock price variance?
2. Is there a linkage (and possibly a lag) between rising stock price variance and falling stock prices?
3. Do we see differences in stock price variance across stock populations?
4. Do indices provide significant protection against stock price variance?

We have additional data to what is reported here. For collaboration opportunities please direct message us.

Our Observations:

We started measuring stock price variance in May 2022. Since then, the variance of daily price changes are elevated, and have recently plateaued. This means that on any given day, stocks will move more than before, regardless of whether up or down, however the rate of growth in variance is zero, or slightly negative. In other words, stock price variance is no longer increasing.

We have heard a great deal about market liquidity, and have a hypothesis that changes in liquidity impacted stock price variance. However, we do not know whether liquidity improved in December 2022, and that seems questionable.

Stock price variance is consistently and significantly higher for stocks of companies with negative net income (and may include companies with non-US HQ locations) than for companies with positive net income (and solely US HQ locations). We also see larger overall declines in unprofitable company stocks over the past year.

Higher price variance and larger price declines are significant in our datasets between populations of US-listed unprofitable companies than US-Listed and US HQ companies that make a profit.

The riskiness of the SPY (the S&P 500 Index ETF) is elevated, but is slightly better (lower) than an equally weighted set of all profitable stocks. We understand the hypothesis and anecdotal evidence that market capitalization weighted indices provide better risk-adjusted returns. We believe that the $SPY is providing better variance readings than an equally weighted set of profitable stocks because of that capitalization weighting.

Key Takeaways:

1. We can improve the variance of our portfolios by market capitalization weighting our data analysis, although this has a high computational cost and would reduce our data accuracy. You can gain this benefit by investing with the S&P 500 ETF.
2. Invest only in profitable companies with US HQ locations if you want less stock price volatility, all else being equal.
3. We have seen a casual relationship between rising stock price variance and falling stock prices. There is no proof of causality. 

Key Research Question:
Why did stock price variance increase from May 2022 through December 2022? 
We hypothesize that this is caused by multiple factors, including reduced liquidity, increased costs of investing (e.g., rising interest rates), and a lower cost of making speculative bets on US equities.

​

1. US-listed stocks with positive net income.
May 17: 1.4 x 10-4
May 20: 1.4 x 10-4
May 31: 1.4   x 10-4
June 1 : 1.4 x 10-4
June 8:  1.4 x 10-4
June 10: 1.5 x 10-4 (up a tick). Expected return 4.75%; 2,013 stocks (dropped significantly)
Min & max expected return 0.01, 0.15 (rounded to 2 decimal places)
June 24: 1.6 x 10-4 (up a tick)
July 6: 1.6 x 10-4
July 11: 2004 stocks (profitable, removed test for positive cash flow, dropped floor cap to $5M)
July 11: 1.9 x 10-4 (up three ticks)
July 26: 1.9 x 10-4 (2,010 stocks)
​July 27: 1.9 x 10-4 (2,014 stocks)
July 29: 1.9 x 10-4 (2,014 stocks)
August 7: 1.9 x 10-4 (2,020 stocks)
August 11: 1.8 x 10-4 (2,165 stocks)
Aug 12: 1.9 x 10-4
​Aug 19: 1.9 x 10-4
Aug 25: 1.9 x 10-4 (2,141 stocks)
Aug 29: 1.9 x 10-4 (2,156 stocks)
Sept 2: 2.1 x 10-4 (1,722 stocks) - tighter data validation to select higher liquidity stocks*.
Sept 6: 2.0 x 10-4 (1,641 stocks) - even tighter data validation
Sep 11: 2.1 x 10-4 (1,568 stocks) - tightened data validation further (to increase liquidity)
Sep 14: 2.2 x 10-4 (1,564 stocks) (up a tick)
Sept 15: 2.2 x 10-4 (1,564 stocks)
Sept 22: 2.2 x 10-4 (1,554 stocks)
Sept 23: 2.1 x 10-4 (1,267 stocks) - increased minimum market capitalization and minimum daily traded volume.
Sept 27: 2.1 x 10-4 (1,269 stocks)
Sept 30: 2.1 x 10-4 (1,272 stocks) - list of stocks is growing (no change to underlying data validation).
Oct 3: 2.2 x 10-4 (1,275 stocks) - list continues to grow with no changes to settings.
Oct 4: 2.2 x 10-4 (1,286 stocks)
Oct 7: 2.2 x 10-4 (1,275 stocks)
Oct 10: 2.2 x 10-4 (1,276 stocks)
Oct 11: 2.3 x 10-4 (1,328 stocks) - we lowered market capitalization floor to $750M from $1B today.
Oct 12: 2.3 x 10-4 (1,347 stocks)
Oct 13: 2.3 x 10-4 (1,361 stocks)
Oct 14: 2.3 x 10-4 (1,358 stocks)
Oct 17: 2.4 x 10-4 (1,491 stocks) - We lowered market capitalization floor to $500MM from $750MM
Oct 18: 2.4 x 10-4 (1,490 stocks)
Oct 21: 2.4 x 10-4 1,497 stocks
​Oct 28: 2.4 x 10-4 1,508 stocks
Nov 2: 2.4 x 10-4 1,510 stocks - kept market capitalization at $500mm
Nov 4: 2.5 x 10-4 1,530 stocks - we reduced market capitalization floor to $400mm from $500mm
Nov 9: 2.5 x 10-4 1,520 stocks
Nov 10: 2.6 x 10-4 1,529 stocks
Nov 23: 2.6 x 10-4 1,745 stocks - Reduced market capitalization to $100mm
Dec 7: 2.5 x 10-4 1,709 stocks - fewer profitable companies...no change in parameters
Dec 15: 2.5 x 10-4 1,708 stocks
Dec 30: 2.5 x 10-4 1,656 stocks
Jan 5: 2.6 x 10-4 1,450 stocks (reduced market capitalization to $100M and increased daily minimum shares traded to 50k). Likely volume required daily was binding, so raising minimum from 20k shares to 50k shares daily reduced ~200 stocks.
Jan 9: 2.6 x 10-4, 1,449 stocks
Jan 12: 2.6 x 10-4 1,383 stocks (minimum capitalization raised to $500M)
Jan 17: 2.6 x 10-4, 1,386 stocks
Jan 20: 2.5 x 10-4, 1,826 stocks (we loosened up the data validation to allow more stocks.) This 


2. US-listed stocks with negative net income.
May 17: 4.3 x 10-4
​May 20: 4.2 x 10-4
May 31: 4.2 x 10-4
June 1 : 4.2 x 10-4
June 8: 4.3 x 10-4 (it finally went up a tick)
June 10: 4.4 x 10-4 (up another tick)
Expected return: 6.24%
1,501 stocks (increased significantly)
Min & max expected return 0.0109, 0.1454
June 24: 4.7 x 10-4 (up three ticks)
1,505 stocks (negative net income - global companies listed on US stock exchanges):
July 6: 4.8 x 10-4 (up a tick)
Expected return of the full set of 1,505 stocks: 6.11%
July 11: 1,444 stocks (unprofitable, floor cap $5M)
July 11: 5.3 x 10-4 (up five ticks)
July 26: 5.4 x 10-4 (up one tick - 1,440 stocks)
July 27: 5.5 x 10-4 (up one tick - 1,489 stocks)
July 29: 5.4 x 10-4 (down a tick - 1,493 stocks)
August 7: 5.5 x 10-4 (up a tick - 1,492 stocks)
August 11: 5.2 x 10-4 (1,658 stocks)
​Aug 12: 5.2 x 10-4
Aug 19: 5.3 x 10-4
Aug 25: 5.2 x 10-4 (1,672 stocks)
Aug 29: 5.3 x 10-4 (1,664 stocks)
Sept 2: 5.7 x 10-4 (1,197 stocks) - we tightened up the data validation (these are higher liquidity stocks)
Sep 6: Did not run
Sep 11: 6.6 x 10-4 (871 stocks) - We tightened up data validation further (more liquidity required)
Sept 14: 6.1 x 10-4 (455 stocks) - we tightened up data validation (required a $1B market cap minimum).
Sept 15: 6.1 x 10-4 (456 stocks)
Sept 22: 6.0 x 10-4 (429 stocks) - down a tick
Sept 23: 6.0 x 10-4 (423 stocks)
​Sept 27: 6.1 x 10-4 (419 stocks) - Fewer stocks as we maintain an aggressive market cap floor (maybe stocks are falling below)?
Sept 30: 6.2 x 10-4 (430 stocks) - more stocks with no change to underlying data validation.
Oct 3: 6.4 x 10-4 (436 stocks) - more stocks with no changes to settings
Oct 4: 6.5 x 10-4 (447 stocks)
Oct 7: 6.5 x 10-4 (427 stocks) (we raised minimum capitalization to $1B & minimum daily trading volume to 50k shares)
Oct 10: 6.4 x 10-4 (423 stocks)
Oct 11: 6.7 x 10-4 (480 stocks) - we lowered market capitalization floor to $750M from $1B today.
Oct 12: 7.0 x 10-4 (479 stocks)
Oct 13: 7.0 x 10-4 (486 stocks)
Oct 14: 6.9 x 10-4 (478 stocks)
Oct 17: 7.0 x 10-4 (585 stocks) - we lowered market capitalization floor to $500MM from $750MM.
Oct 18: 7.0 x 10-4 (592 stocks)
Oct 21: 7.1 x 10-4 581 stocks
Oct 28: 7.2 x 10-4 611 stocks
Nov 2: 7.3 x 10-4 657 stocks - lowered market capitalization floor to $400mm from $500mm
Nov 4: 7.3 x 10-4 662 stocks - no change (but I know of at least one stock that moved from profitable to unprofitable)
​Nov 9: 7.3 x 10-4 666 stocks
Nov 10: 7.7 x 10-4 676 stocks
Nov 23: 8.0 x 10-4 833 stocks - lowered market capitalization to $100 mm
Dec 7: 8.0 x 10-4 805 stocks - fewer unprofitable stocks, no change to parameters. Must be getting less valuable or delisted...
Dec 15: 7.9 x 10-4 751 stocks (We removed several biotech stocks from future runs. It is our intention to remove them all.)
Dec 30: 7.7 x 10-4, 643 stocks (increased minimum market capitalization limit which reduced the number of stocks)
Jan 5: 7.9 x 10-4, 689 stocks (reduced minimum market capitalization to $100M and increased minimum shares traded daily to 50k). Likely market capitalization parameter was binding (so a lower market cap allows more stocks)
Jan 9, 7.8 x 10-4, 682 stocks (excluding most therapeutics, biotechnology and drug discovery firms)
Jan 12: 7.5 x 10-4 534 stocks (raised minimum market capitalization to $500M and removed more biotech stocks)
Jan 17: 7.5 x 10-4 536 stocks
Jan 20: 6.6 x 10-4, 936 stocks (loosened the restrictions on data validation to include more stocks).

3. $SPY, or the S&P 500 Index ETF:
May 17: 1.2 x 10-4
​May 20: 1.3 x 10-4
May 31: 1.3 x 10-4
June 1 : 1.3 x 10-4
June 8:  1.3 x 10-4
June 10: 1.4 x 10-4 (up a tick)
​June 24: 1.6 x 10-4 (up two ticks)
2,040 stocks (US companies, US listed, with positive net income and positive CFFO):
July 6: 1.6 x 10-4
Expected return of the full set of 2,040 tickers: 4.75%
July 11: 1.6 x 10-4
July 26: 1.6 x 10-4
July 27: 1.6 x 10-4
July 29: 1.7 x 10-4 (up a tick)
August 7: 1.7 x 10-4
Aug 10: 1.7 x 10-4
Aug 12: 1.7 x 10-4
Aug 19: 1.7 x 10-4
Aug 25: 1.7 x 10-4
Aug 29: 1.8 x 10-4 (up a tick)
Sept 2: 1.8 x 10-4
Sept 6: 1.8 x 10-4
Sept 11: 1.8 x 10-4
Sept 14: 1.9 x 10-4 (up a tick)
Sept 15: 1.9 x 10-4
​Sept 22: 1.9 x 10-4
Sept 23: 1.9 x 10-4
Sept 27: 1.9 x 10-4
Sept 30: 1.9 x 10-4
Oct 3: 1.9 x 10-4
​Oct 4: 2.0 x 10-4 (up a tick)
Oct 7: 2.0 x 10-4
Oct 10: 2.0 x 10-4
Oct 11: 2.0 x 10-4
Oct 12: 2.0 x 10-4
Oct 13: 2.0 x 10-4
​Oct 14: 2.0 x 10-4
Oct 17: 2.0 x 10-4
​Oct 18: 2.0 x 10-4
​Oct 21: 2.1 x 10-4
Oct 28: 2.1 x 10-4
​Nov 2: 2.1 x 10-4
Nov 9: 2.2 x 10-4
Nov 10: 2.3 x 10-4
Nov 23: 2.3 x 10-4
Dec 7: 2.3 x 10-4
Dec 15: 2.3 x 10-4
Dec 30: 2.3 x 10-4
Jan 5: 2.5 x 10-4
Jan 9: 2.3 x 10-4
Jan 12: 2.3 x 10-4
​Jan 17: 2.3 x 10-4
Jan 20: 2.3 x 10-4

4. $QQQ, or the NASDAQ Composite 100 ETF:

Nov 4: 3.8 x 10-4 (higher than I expected, higher than both all profitable stocks and $SPY)
Nov 9: 3.8 x 10-4
Nov 10: 4.0 x 10-4
​Nov 23: 4.0 x 10-4
​Dec 7: 4.1 x 10-4
Dec 15: 4.1 x 10-4
​Dec 30: Not measured
Jan 5: 4.1 x 10-4
​Jan 9: 4.1 x 10-4
​Jan 12: 4.1 x 10-4
​Jan 17: 4.1 x 10-4
Jan 20: 4.1 x 10-4

5. All stocks, up and down, profitable and unprofitable, running together:

Jan 20: 3.6 x 10-4 2,836 stocks (profitable and unprofitable stocks, including very low market capitalization stocks)

​Assumptions: portfolios are evenly weighted, for 253 trading days, using adjusted closing stock prices.

Note (*): We continue to adjust our data validation parameters to help us select stocks that can be readily purchased and sold for our Chicago Quantum Net Score model portfolio / managed accounts / hedged fund.

Daily fluctuations in the number of stocks is likely due to changes in data validation parameters, followed by changes in earnings.

Thank you Finviz, Yahoo Finance and Intrinio.