The “Magnificent Seven” has been a dominant force in the market for years, consistently carrying the cap-weighted indices higher while equal-weighted schemas languished. The last few months have seen that trend change as the S&P 500 Equal Weight Index (SPXEWI) outpaces the S&P 500 Index (SPX) by roughly 3.5% year-to-date (2.64% versus -0.94%). The Magnificent Seven stocks, a large portion of the SPX, have really struggled this year relative to the SPXEWI. Using the Roundhill Magnificent Seven ETF (MAGS) as a representative, MAGS is down more than 6% this year with every Magnificent Seven stock in the red year-to-date. However, over the last year, MAGS is up 33.67% versus 13.17% for the SPXEWI and 20.78% for the SPX. So, we’ll need to see a longer or more severe period of underperformance before the evidence starts to pile up against the Magnificent Seven. One interesting point is the dispersion within the Mag Seven stocks over the last year with Alphabet (GOOGL), NVIDIA (NVDA), and Tesla (TSLA) each gaining over 70% while Meta (META) and Microsoft (MSFT) returned less than 10%.

Focusing on MAGS’s chart action, the fund broke a quadruple bottom in February, its first sell signal since April 2025. The immediate concern at first glance is the lack of established support nearby. Prior resistance at $58 may act as support but below that there is no established support until $46.50. The fund score for MAGS is still in acceptable territory at 3.63, but it does have a 2.21 negative score direction, which speaks to its decline in relative strength over the last few months. While these handful of mega cap names have slowed down, many are still in strong standing on an individual basis. When looking to trim or exit one of the Mag Seven names that are likely in at least a few of your clients’ portfolios, be sure to check on the individual technical pictures for each even if the overall group has slowed down. Secondly, the strong performance of these handful of names may have led portfolios to having an outsized allocation. So, this could be a good opportunity to reallocate portfolios towards emerging areas of relative strength without completely selling out of the Mag Seven. If these mega caps continue to weaken, then more changes will be needed.

Lots of recent research has hit on changes around the platform and what that could ultimately mean for the state of trendfollowing within your portfolio. Growth names have rolled over slightly and the uptick in energy focused commodities have led to several shifts within the broader DALI and Asset Class Group Scores rankings. As strength shifts, we thought it would be useful to look underneath the hood of our model builder tool… to detail how you can utilize the platform to create bespoke custom models to take advantage of shifts of strength. More specifically, we will use the model builder to discuss the different levers you can pull… from Technical Attribute Score overlays, sector maximums, etc. and the overall effect on performance, risk, and model compatibility.

To lay the groundwork, we will utilize a matrix based model focusing on the NDX as our starting universe. From there, we will look to own the top five stocks within the overall rankings based on their buy count, evaluated on a monthly basis. The model will sell once the holding falls out of the top third of the rankings. This setup is similar to the pre-made NDW NDX Model (NDXMODEL.TR) except that model focuses on ten holdings and cuts our losers a bit shorter than the premade. Enter the first benefit of the custom modeler- the ability for you to put your own spin on an existing NDW strategy, without compromising the integrity of the overall process. This model on its own is quite impressive- growing at a CAGR of ~18.50% since 2001 vs. the ~9.65% for benchmark NDX. For the uninitiated- this screen certainly has a lot going on. A few important places the analysts look at when we create models are the returns (obviously) across different timeframes & the max 6-month drawdown and turnover numbers. We want our drawdown to be low (in this case this model’s drawdown is slightly worse than our benchmark) & turnover metrics to be “reasonable”. That definition is different for every client… but most of our premade models hover around 100%-200% per year. In this case, our model is above that threshold- turning over about 2.3 times per year. This works for some clients, not for others. Another important thing to keep in mind is the type of account you place these models in- a qualified account can withstand more trading than something that your clients will be paying taxes on in the near-term.

This is the most “basic” model possible: we gave the system a target number of holdings, buy and sell criteria, and an evaluation frequency. There are, however, several different levers you can pull to “enhance” the model in one form or another. The first adjustment we can make is applying an overlay filter that instructs our model to buy those stocks at the top of the matrix…. But only if they are rated a 4 or 5 technical attribute score. In doing so, we can add another layer to our security selection process. Adding this in adds an extra ~700% to our cumulative returns (+~.4% p/year) since December of 2000. Perhaps most useful about this addition is that it decreases our max 6-month drawdown by roughly 10% to -40.13%, a metric lower than that of our benchmark. When you think about it, this makes sense… by ensuring that there are only 4 or 5 rated stocks in our portfolio, we avoid the 3’s… which become 2’s, 1’s, or 0’s. Our turnover increases slightly, but not enough that the decreased drawdown isn’t “worth it”.

While today’s piece focused on a possible strategy in the model builder, know that this process is applicable to those of you not using the builder as well. By isolating those stocks at the top of the matrix while ensuring names that also score well according to the TA system, we can add more weight of the evidence to our picks. While we cant ensure that this strategy will be additive in all scenarios, those of you looking to test different strategies may find it useful to play around with different overlays. We didn’t discuss cash triggers which allow your strategies to raise varying levels of cash for more defensive clients, or sector maximums to ensure that you don’t get too overallocated to any one specific area of the market. It is also worth noting that a frequent use case for many is to blend together different models (a combination of 5 holding models for various universes like NDX, SPX, .DJIA, or another custom sleeve) to create a more robust 15-20 holding model that satisfies compliance requirements.

All this to say, there are a virtually endless number of possibilities available to you to play around with in the builder. If you would like to discuss building out a model similar to this, feel free to email at miles.clark@nasdaq.com for more information or a trial of the model builder.

Each week the analysts at NDW review and comment on all major asset classes in the global markets. Shown below is the summary or snapshot of the primary technical indicators we follow for multiple areas. Should there be changes mid-week we will certainly bring these to your attention via the report.

Remember, these are technical comments only. Just as you must be aware of fundamental data for the stocks we recommend based on technical criteria in the report, so too must you be aware of important data regarding delivery, market moving government releases, and other factors that may influence commodity pricing. We try to limit our technical comments to the most actively traded contracts in advance of delivery, but some contracts trade actively right up to delivery while others taper off well in advance. Be sure you check your dates before trading these contracts. For questions regarding this section or additional coverage of commodities email james.west@nasdaq.com.

Data represented in the table below is through 3/11/26:

Portfolio View - Commodity Indices

Cryptocurrency Update

Cryptocurrency Video (2:59)

Daily Option Ideas for March 11, 2026