Jeffrey Cohen, President US Advanced Computing Infrastructure, Inc.  April 12, 2021

This is our first BLOG Post where we discuss some of the learnings from Q1 2021 as relates to our 'short' or 'reverse' model.  We call it the reverse model because we found a way to select the worst, or most inefficient portfolios that can be held by an investor relative to the benchmark (either SPY or QQQ at this point).

Why would anyone want an inefficient stock portfolio?  A portfolio that maximizes variance and minimizes expected returns.  This portfolio provides stock price movement (noise) without clear market direction.  The expression "all hat and no cattle" comes to mind. These are a portfolio of stocks to avoid, not to buy and hold...to short if you invest that way to balance or otherwise manage your market exposure.

We have been running this type of model for some time now, and we ran this job on Friday, April 9, 2021 market close data.  Today was the first day we dug into the results in detail.  We picked 4 of the 5 industry groups and did a cursory look at each stock, how it moved in the past year, and how they earn their revenues.  We did not go into detail into financial statements, balance sheets, legal proceedings, or otherwise research the stock.  We did not trade this data, but we began the research we would do to feel comfortable trading these stocks. This level of work took about half a day (for ~35 of the 50 stocks) and we learned a great deal.

We also created a portfolio of the 50 worst, or most inefficient stocks, which we call 'dogstars' in contract to 'allstars.'  In the first day, 41 fell (and many fell double digit % losses), 2 were flat, and 7 rose (at most single digit % gains).  The market was down overall, which is likely a contributing factor (market falls, dogstars fall more). 

Industry groups:  The top industry group represented was pharma/bio-science.  Other industry sectors well represented were retail / consumer discretionary products, oil & gas, and technology (software and services).


In conclusion, we picked portfolios of stocks to avoid for a client, and we took the 50 'most inefficient' or 'worst' stocks to hold in a portfolio and did additional analysis on them.  Those stocks were chosen by the Chicago Quantum Net Score, a quantum algorithm, based on April 9, 2021 market close data (going back one year).  On the first day of trading, starting with market open and ending at market close, 41 of the 50 stocks fell (and more significantly than the 7 that rose...and 2 were flat).  The overall market was down for the first day in a while. 

We are curious to see if this trend continues tomorrow (*and how long it continues). 

​Disclosure:  We are not financial advisors and this is not financial advice.  We will not share details of the stocks chosen as this run was paid for by a client.  We do not have a position in any of the stocks included in the 50 stock portfolio (long or short), but may begin to take short positions in the future.  We are long SMLP and NGL at this time, other than our fully passive index and fund investments. 
Please do your own due diligence before you invest.

Chicago Quantum (SM) is a protected service mark of US Advanced Computing Infrastructure Inc.

This year has seen a significant investment in our technology infrastructure.  

​In Q1 2021, we replaced our iMac 2013 with an iMAC Pro 2017.  This gives us significantly more processing power and graphics capability to do our daily work (administrative, marketing and finance), as well as easily and efficiently run our smaller stock picking jobs.

In Q2, we purchased an upgrade to max the RAM of our HP server (server 1) which was purchased in 2013.  This server is becoming insufficient to run our larger analytical jobs, as we have added more stocks, new steps in data validation (including new proprietary solvers), and new analytics we perform on stocks.  It just runs a little too slow.

In Q2 we purchased a new Dell server, along with significant upgrades that are being received and installed this week.  This server is significantly faster and more capable than Server 1.  As an example, the CPU was first released in 2019, has a 22% faster base clockspeed, and this system has 4x the RAM, and runs ~80% faster, than server 1.  This new server made its debut on April 10, 2021, and has already provided a better (deeper) solution from our production run this past weekend.

What's next? 

We are currently in discussions with market data services providers (early stage - new bidders welcome) to purchase data feeds on stock and potentially other financial instruments.

We are also considering whether to expand the payroll...so stay tuned.

Regards,
Jeffrey Cohen
President, US Advanced Computing Infrastructure, Inc.
April 12, 2021

It was fun this week.  I had prepared topics and it went pretty smoothly.  Interesting days in the markets, and good progress within the firm.

https://youtu.be/ObDA0CofsAk

​

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

"It was a dark and stormy night..."

NASDAQ was falling, long bond yields were screaming higher, the dollar was strengthening, and fintwit tears could have overflowed Lake Michigan.  People were scared of the market.  Bitcoin had fallen $10k per coin, and the meme stock aftermath (GME and AMC) was a shattered set of dreams for retail and hedgies alike. The other thing we noticed was that annual returns for our three market exchanges we use in our model were all over 20%, and one was 65%.  These are heady returns.

In other words, it was time for one of our clients to think about a balance of long and short positions.  Get to the safe spot where you make money purely on Alpha and not on Beta.  Check out our tab 'Market Questions & Data' for explanation of terms.

We were ready for the challenge and revised our original Chicago Quantum Net Score model (our quantum algorithm) to pick 'dogstars' or stocks that had exceptional volatility, without a sense of direction.  In other words, their volatility exceeded their expected market returns.

So, we did it.  We are now offering that service on our website for clients that want a set of stocks to look at not owning (or shorting, or buying puts).  So far, the results are pretty good.  Initially we see the stocks falling, but then recoveries happen.  We plan to watch these stocks for a few months and report on progress.

It did find a stock that is in real trouble...although we had to study the financials and bond covenants and amended agreements to find out how badly they were in pain.  They owe more than $80M to a sharp investment house in Jersey (no, not New Jersey) due March 31, 2021.  I cannot see a way out for them...and so the model has already had value for our client.

Check it out for yourself.  A big run is $750 (where we evaluate exchanges) and a small run (where you provide the tickers) is $150.  Until supplies last, we will buy and donate a box of Girl Scout cookies for each paid run.

You can call me, and I can walk you through the logic and math.  Contact info is in the 'Contact' tab.  This is pretty cool work.
​