Technical attributes can be an efficient way to communicate technical strength for stocks, but how well do they hold up?
There are many different technical data points available on the NDW platform, but one of the most popular continues to be our technical attribute ratings for stocks. We have found these to be a great way to aggregate several different technical readings into one easily understood rating. The technical attributes have also been easier for the uninitiated to understand; potentially helping client conversations about positions with emotional strings attached, or with initial client prospecting efforts.
This all sounds good, but how well does it really hold up? A few years ago, we looked to answer that question by revisiting the entire attribute rating system to stress test different use cases for the attributes. Some highlights from that whitepaper are included below, with the data recently updated through the end of 2025. All of our whitepapers can be accessed here.
Introduction
Over the years, Nasdaq Dorsey Wright has created many innovative technical indicators based on momentum using Point and Figure charting. One of our most popular indicators in this space is the “Technical Attribute” rating we apply to each stock. The rating ranges from 0 (lowest strength) to 5 (highest strength) with values of 3 or higher, generally regarded as investable.
The purpose of this study was to determine how effective technical attributes are from a portfolio management perspective. If we “buy” a portfolio of high attribute stocks, do we outperform the market? Alternatively, what happens if we buy a portfolio of low attribute stocks? We found that high attribute portfolios had a strong propensity to outperform, with the largest outperformance reserved for the highest ratings (5 attribute stocks), while low attribute portfolios had a marked tendency to underperform under most market conditions.
We also took this initial insight and progressively refined the concept to create a portfolio with good performance, manageable portfolio size, reasonable turnover, and simple to understand portfolio construction process. The result is a portfolio which buys 5 attribute stocks ranked highly in Nasdaq Dorsey Wright’s matrix system (explained below) and waits to sell them when their attribute rankings fall to 2 or lower.
Studies
To determine how technical attributes work in a portfolio management context we first calculated technical attributes for the top 1000 market cap stocks every year from 1992-2025. We chose the top 1000 market cap stocks as they’re highly liquid and thus more practical for the average financial advisor to buy. Next, we ran five separate tests to analyze the properties of the rankings and their applicability to a hypothetical client’s portfolio. Each test provides insights which are then used to inform the following test until we end up (in Test #5) with the portfolio mentioned in the introduction of this paper. Unless otherwise specified, all tests were conducted between 1992 and 2025, utilizing the top 1000 market cap universe. They also buy equal weight positions in all stocks matching the test criteria, rebalance/reconstitute monthly, and update the top 1000 market cap universe at the end of every year.
Test #1 – 3’s and Higher vs. 2’s and Lower
For the first test, we chose to “buy” all stocks with 3 or more attributes in one portfolio and compare them to a portfolio owning all stocks with 2 or fewer attributes. The idea is to test whether high attributes perform better than low attributes in general.
As the graph shows, the “3’s and Higher” portfolio outperforms the “2’s and Lower” portfolio by 2.50%/year. It also outperforms the S&P 500 and an equal weighted index of the top 1000 market cap stocks (Equal Weight 1000 in the graph) while the “2’s and Lower” portfolio underperforms both benchmarks during the study. Interestingly, the typical performance difference is more extreme than shown as, with the benefit of hindsight, low attribute stocks tend to do very well after major bear markets, which makes their returns look better over the entire period than what you may get by holding them for any particular year. If we remove 2003 and 2009 from the analysis (both included the initial moves up off a major bear market) and look at the difference in returns for all other years we find that the average outperformance moves up to 3.53%/year from 2.50%/year. Of course, you can’t just ignore the performance that doesn’t help you, but it is an interesting data point. We also found that the volatility of the “2’s and Lower” portfolio is much higher than the “3’s and Higher” portfolio while offering lower returns (meaning its Sharpe ratio is lower as well).
While this test illustrated that higher attributes seem to offer better performance, the next test seeks to understand whether the additional performance is distributed equally amongst the high attribute rankings or whether a particular ranking fares better than others, which can be found in the full white paper.
