Pre-election, Nov 1 analysis: Reduced Volatility, Enhanced Return US common stock portfolio

This is a short article. On Nov 1 we downloaded and validated 3,206 stocks. All are US common stocks from NYSE, NASDAQ and NYSE American (tape B). From that, we chose our first portfolio.

Update: Nov 25, 2020 to reflect learnings on BETA values and the reason (possibly) on why our model did not pick a smaller portfolio.

Portfolio 1: 135 stocks

['AAPL', 'ABT', 'ADBE', 'AJRD', 'AKAM', 'AMD', 'AMH', 'AMP', 'AMZN', 'ANSS', 'AOD', 'ATUS', 'ATVI', 'AVLR', 'AWK', 'BA', 'BAH', 'BAX', 'BCE', 'BIIB', 'BIOX', 'BSX', 'CB', 'CDNS', 'CHD', 'CHL', 'CI', 'CL', 'CLVS', 'CMS', 'CNS', 'COUP', 'CR', 'CRM', 'CRUS', 'CSII', 'CTXS', 'DCPH', 'DOV', 'EBR', 'ECL', 'ENBL', 'EQIX', 'ETV', 'ETX', 'EVRG', 'EVT', 'EXD', 'EXG', 'EXPO', 'FAST', 'FB', 'FCX', 'FE', 'FELE', 'FISV', 'FR', 'GL', 'GLW', 'GNTX', 'GOOGL', 'GPN', 'GSKY', 'HCA', 'HD', 'HE', 'HELE', 'HON', 'HTY', 'HY', 'IAC', 'IBN', 'ICE', 'IDXX', 'ILMN', 'IMMR', 'INTU', 'IT', 'KLAC', 'KPTI', 'LRCX', 'LUV', 'MAIN', 'MELI', 'MGRC', 'MKC', 'MMS', 'MRK', 'MSFT', 'MSGS', 'MSI', 'MTSI', 'NDAQ', 'NGVT', 'NOW', 'NTRS', 'NVDA', 'NVS', 'ORCL', 'OTTR', 'PBF', 'PDCE', 'PEG', 'PEP', 'PG', 'PINS', 'PLAG', 'PLAN', 'PLD', 'PNC', 'PXD', 'QCOM', 'RGA', 'RIO', 'RMD', 'SCL', 'SCPH', 'SEIC', 'SJM', 'SNPS', 'TEF', 'TGT', 'TRS', 'TSLA', 'UFS', 'USB', 'VNE', 'VNRX', 'VRSK', 'VRTX', 'WRN', 'XLNX', 'ZBRA', 'ZTS', 'ZUO'] 135.0

Out of those 135 stocks we chose a 2nd portfolio.

Portfolio 2: 68 stocks:

GA Seeded     -0.000003       0248 
['AAPL', 'ABT', 'ADBE', 'AJRD', 'AKAM', 'AMD', 'AMZN', 'ANSS', 'ATUS', 'ATVI', 'AWK', 'BA', 'BAX', 'BIIB', 'CHD', 'CI', 'CR', 'CRM', 'CTXS', 'ECL', 'ENBL', 'ETV', 'EVT', 'EXD', 'EXG', 'FAST', 'FB', 'FCX', 'FELE', 'FISV', 'GL', 'GLW', 'GOOGL', 'HCA', 'HD', 'HE', 'HON', 'IBN', 'IDXX', 'ILMN', 'IMMR', 'INTU', 'KLAC', 'KPTI', 'LUV', 'MKC', 'MMS', 'MSFT', 'MTSI', 'NDAQ', 'NTRS', 'NVDA', 'ORCL', 'PBF', 'PEG', 'PEP', 'PG', 'PLAN', 'QCOM', 'RIO', 'SEIC', 'SJM', 'SNPS', 'TEF', 'USB', 'VNE', 'VRSK', 'ZTS'] 68.0

Why does our best portfolio contain 68 stocks and not fewer? We can only assume the data suggests a flight to the safety of a more diversified portfolio.

Update: We have seen market declines change the results of our model. I recall running our models in March 2020 during the COVID 19 collapse. This is one contributing factor.

Update: A second factor was that changed our CQNS formulation during a short period of time in a few of our solvers. We had increased the expected returns ‘too much’ in our solvers, and as a result the model tilted towards reducing volatility to the extreme. We saw low-BETA stocks chosen, and an increase in the number of stocks chosen to maximize diversification. In short, we engineered a flight to safety by changing our CQNS formulation (see the sign below). We have since reversed that change, and gone back to our QUBO-friendly CQNS formulation that was first introduced in our 40 stock paper.

Image for post
Image for post
Photo by Skitterphoto by Pexel, warning about slipping and falling

We found a few changes this week versus prior runs:

Actual riskfree rate =   0.56%
Use floor riskfree rate = 1.00%
Actual GSPC return = 0.09733521403858823
Use actual S&P500 rate = 9.73%
Actual RUT return = -0.0005966791567518978
Use floor Russell 2000 rate = 2.50%
Actual NASDAQ return = 0.3751259234091236
Use ceiling NASDAQ Composite rate = 20.00%
===========================================
Market return = 9.74%
Image for post
Image for post
Updated: November 19, 2020 market returns used to set expected returns.

2. We substitute the ‘SPY’ S&P 500 ETF for the S&P 500 Index to allow non-institutional clients to buy both the portfolio stocks and the index. Using ‘SPY’ does reduce BETA from a possible range of (0.01,4) to (0.763,1.22). This reduces expected market returns. However, we also reduce our CQNS_Power setting to set Variance = Expected Return for the 135 stock portfolio, so this should mitigate the effect of the change in index.

3. The expected market return for 135 stocks has a significantly lower Sharpe ratio of 1.60. The standard deviation of the portfolio is significantly higher than before.

ALL ASSET PORTFOLIO - equally weighted 
Expected Return = 10.85%
Expected Return cubed - true 0.0046161
Variance = 0.0046161
Standard Deviation = 6.79%
Chicago Quantum Net Score -2.6021e-18 (set to zero)
Chicago Quantum Ratio = 0.9784
Sharpe Ratio = 1.60

Updated: We found a matching D-Wave Systems portfolio this morning (64 stocks out of 135), with a valid solution using our ‘true’ CQNS score.

Our problem embedded successfully (2 of 5 times), with the last time using a higher than normal chain strength of 2. Run times this morning varied from 4 to 12 minutes of waiting time.

Next Steps:

Each time the market shifts, we learn more. We will work through a slight formulation change to improve the power of our QUBO solvers, consider whether ‘SPY’ is the correct market index vs. QQQ, and consider whether to use the new D-Wave Systems Advantage (TM) system with fewer than 134+ stocks to improve quantum performance. Stay tuned for more updates.

We are beginning to code our ‘higher moments’ of stock prices new formulation that could be interesting for options pricing.

Updated: Our final learning is that when we see market performance of our portfolio choices that run counter to our ‘enhanced return, reduced volatility’ thesis, we dig deeper and often learn more. In this case, we have greater confidence in our CQNS formulation, and have taught our solvers to go deeper and find better solutions for our clients.

Have a great week trading before the US Presidential election.

Jeffrey Cohen, President, US Advanced Computing Infrastructure, Inc.

Jeffrey Cohen, President of US Advanced Computing Infrastructure, Inc., & founder of Chicago Quantum (SM). We use quantum algorithms & our quantum platform

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store