I’ve long been a fan of Zacks, the investment research company that brought quality analysis and analytical tools to the trading public independent of the brokerages. I’m still a fan, but in setting up a trading system, I’ve uncovered a few items that make me wish that they would expand their tool set just a little.
I’ve been using Zacks to set up a stock trading system. Since the Fed dropped its funds rate below 2% last September, I’ve lost interesting in government bills and notes and vehicles to house my cash reserves.
I’ve turned to stocks, and as regular readers will know, I’ve been trying to work out a system.
The company was founded in 1978 on the premise that earnings analysis in aggregate by those brokerage analysts was the best way to see ahead to a company’s prospects, earnings being the ultimate basis of stock prices and brokerage analysts being important in setting trader expectations for the future of those prices.
Not exactly the stuff of quants programming their split-second AI underlings, but an approach that has proven solid for many decades.
For me, the beauty of Zacks lies in its analytical tools, a database of information related to 4,000-plus stocks, including fundamentals and analyst opinions, that I can write queries against and download into a spreadsheet for further analysis, all for the price of $20.75 a month. Not out of reach for a trader with limited capital.
My basic outline for a set of trading rules goes like this:
- Rule for the size of the portfolio, in either value or number of symbols (portfolio sizing).
- Rule for determining the size of each trade (trade sizing)
- Rule for conditions that trigger entry (buy signal).
- Rule for how long to hold before first evaluating (holding period).
- Rule for how often to evaluate the stocks in a portfolio (evaluation period).
- Rule for conditions that trigger exit (sell signal).
Zacks comes into play with Rule 3, when to enter. Using the company’s database and my spreadsheet analysis, I can set the conditions for entering a trade — so much change in analyst evaluations, for example, or so much acceleration in momentum, and lot more. I generally aim for a poll of 25 or 30 stocks from the Zacks query and then pick the trade using my spreadsheet analysis.
When a stock no longer appears on the query, then that’s the signal to sell, without second guessing or remorse.
Zacks’ educational materials recommend slowing down process by evaluating positions and trading once a month or for the obsessive, once a week. I’m used to trading once a day, which shows my level of obsession.
Another brake recommended by Zacks is to not exit a position for the first four weeks after entering it, to avoid whipsaws.
It seems to me to slow down trading through something as arbitrary as a period of days. Back the old days (last spring) it made sense to slow trading, because each trade generated a commission that cut into profits and enhanced losses. But since then many of the major brokerages, following the example of the upstart Robinhood, have eliminated commissions on stock trades.
That simple act has created a whole world where we can, for the first time in my lifetime, trade as often as we wish without a penalty for trading.
Commissions also required that we set a minimum size on trades, in order to reduce the percentage of our profits gobbled up by commissions. In a commission without commissions, there’s no penalty for keeping positions small, thereby increasing diversification.
So if a sell signal invalidates a position, as it does under a system-based strategy, why would I want to hold an invalid trade for a month or a week or even a single day? Either a trading system deserves trust, which argues for a quick response to signals, or it can’t be trusted, which argues for changing to a new system.
My second dissatisfaction with Zacks has to do with the ranking system and whipsaws.
The company’s ranking system — 1 for a strong buy recommendation, 2 for buy, 3 for hold, 4 for sell and 5 for strong sell — is relative. That is, 1 and 5 each account for 5% for the pool, 2 and 4 for 15% each, and 3 for 60%. If a ranking is on the margin, it can change very quickly.
I’ve not found a way to get the score that underlies each Zacks rank. If a given signal has a rank of 1, the top 5%, it makes difference to know wether that’s the top 4.99% or the top 1%. Their query system does allow me determine how much the rank has changed in the past four weeks, but nothing further.
The remedy, of course, would be to increase the granularity in the Zacks score. Perhaps 1A for the top 1% down to 1E for just above the boundary.
So that’s what I’ve been working on for the past couple of months. On Friday I did a restructuring into two portfolios, Momentum (which is prone to whipsaws), and Analyst Revisions (which seems to be less prone to whipsaws).
I also want to create an Income portfolio based on Zacks, rather than the no-system portfolio I have today.
And a third possibility that I’m considering is to analyze solely based on the Zacks rank, Rank Change over the last four weeks, and then limit the choices further by market capitalization or some other characteristic.
Here’s a list of the stocks in each portfolio as of Friday, six symbols in each:
Momentum: CROX, CUE, NAVI, NSIT, ORN, SSD
CUE and NAVI dropped off the screen for action on Monday, giving sell signals, so I shall sell them and based on new analysis, buy AXE and MUSA.
Revisions: AUY, CNNE, FAF, IBP, MHO and UCTT.
No action required.
By Tim Bovee, Portland, Oregon, November 16, 2019
Read More »