When Adding Another Stock Makes Your Portfolio Less Diversified
September 7, 2026
Altsets
Research by Altsets Research
Test the next stock by the economic paths it adds to the portfolio, not just its ticker or sector, because a new position can reduce issuer risk while increasing shared customer or bottleneck exposure.
Data used:Altsets Supply Chain Intelligence: 90k+ entities, 400k+ relationships, 20+ years of history.
Key findings
- The supplied Altsets network connects both Micron and SK Hynix to Nvidia, while current company announcements tie both memory suppliers to Nvidia's next-generation AI roadmap.
- Adding a second Nvidia-linked memory supplier can improve issuer diversification while doing much less for customer diversification, so marginal network overlap should be considered before the trade rather than after a shared catalyst appears.
Adding another stock does not always make a portfolio more diversified. Sometimes the new position adds a different ticker, a different management team, and a different set of financial statements while increasing exposure to the same customer, supplier, bottleneck, or product cycle already sitting underneath the portfolio.
Micron and SK Hynix make the idea easy to see. The supplied Altsets network connects both memory companies to Nvidia as a customer, and current public evidence ties both companies directly to Nvidia's next generation AI-memory roadmap. Micron says its HBM4 is in high-volume production for Nvidia Vera Rubin, while SK Hynix has announced a multi-year technology partnership with Nvidia covering next-generation memory for AI factories. An investor who already owns one of those memory suppliers should not assume that buying the other creates a fully independent AI bet.
The right question is what the new stock adds underneath the ticker
Most portfolio construction starts with what the candidate company is. Is it a semiconductor stock, a software stock, a bank, or an industrial? Dependency-aware investing asks a different question before the trade: what new economic paths does this company add, and which existing paths does it repeat?
That distinction matters because a portfolio can become broader by ticker count while becoming narrower around one external company. A second stock can reduce company-specific risk while increasing customer concentration at the portfolio level.
Micron and SK Hynix can diversify issuer risk without diversifying Nvidia demand
Micron and SK Hynix are different businesses with different facilities, balance sheets, product mixes, management teams, and competitive positions. Owning both can therefore reduce the risk that one company executes poorly.
At the same time, both companies have current commercial relevance to Nvidia's AI platform. If the investor's portfolio already depends heavily on Nvidia-related AI infrastructure demand, adding the second memory company can deepen that shared dependency instead of adding a new source of economic independence.
That is not an argument against owning both. It is a reason to describe the trade correctly.
Marginal diversification should be measured before the order is placed
The useful comparison is not whether the new stock is diversified in isolation. The investor should compare the candidate's important customers, suppliers, and bottlenecks with the dependency map of the portfolio that already exists.
A stock with several concentrated relationships can still be an excellent diversifier if those relationships are new to the portfolio. A company with a broad customer base can add very little diversification if its most important outside nodes are already repeated across several holdings.
This makes diversification a marginal decision. The same stock can be a strong diversifier for one investor and a poor diversifier for another.
Sector diversification can move in the wrong direction
Imagine an investor owns a memory supplier and decides to diversify into another company with a different ticker and different home market. A conventional screen may show more geographic and issuer diversity after the trade.
If both companies rely on the same AI customer and the same product cycle, however, the portfolio has added another claim on a shared demand source. The labels became more diverse while the underlying economic path became more concentrated.
This is why supply-chain data should complement sector and country allocation rather than replace them. Each layer answers a different question.
A new stock should earn its place by adding a new path or a better version of an old one
Repeated dependencies are not automatically bad. An investor may deliberately want more exposure to Nvidia's AI ecosystem because that is the core thesis. In that case the second stock can still make sense if its valuation, competitive position, upside, or risk profile offers a better way to express the theme.
The difference is intentionality. The investor knows the new position is increasing exposure to an existing dependency and can size the trade accordingly.
If the goal is diversification, the standard should be higher. The new stock should introduce meaningful customers, suppliers, geographies, or bottlenecks that are not already dominating the portfolio.
This changes how watchlists should be ranked
A normal watchlist might rank stocks by expected return, valuation, or conviction. A dependency-aware watchlist can add one more column: incremental network overlap with the current portfolio.
That does not require a single composite score. The investor can simply ask which important external nodes would be repeated if the stock were added and whether those repeated nodes already represent large positions or known catalysts elsewhere in the account.
A candidate that looks slightly less exciting on a standalone basis can become more attractive if it adds a genuinely independent economic path.
The conclusion is about marginal independence
The number of holdings in a portfolio is not the same as the number of independent bets. The relevant diversification question is whether the next stock gives the portfolio another way to make money or simply places more capital behind the same outside companies and product cycles.
For an investor already exposed to Nvidia-linked AI memory, adding another Nvidia-linked memory supplier may improve issuer diversification while doing much less for customer diversification. That is the conclusion the relationship map makes visible before the trade rather than after a shared catalyst hits both positions.
The supply-chain diversification guide explains why sector and geography do not capture every dependency. The shared-customer guide shows how common downstream nodes can be found across existing holdings.
For relationship definitions and evidence limits, read the Altsets methodology.