Elon Musk's bid to combine SpaceX and Tesla might be good for Musk. It looks less so for shareholders -- especially Tesla's.

A combination would bring under one roof Musk's main business pursuits, streamlining the sprawling empire of the world's richest person.

On the plus side, the resulting conglomerate's massive size and diverse revenue streams could help drive down its cost of capital, making it easier for Musk to finance the multiplanetary sci-fi future he envisions.

But on the other hand, combining everything from satellites to social media, robotics, artificial intelligence, solar power and electric vehicles risks creating a confusing mess of a company with little uniting it besides Musk's own imagination. A hefty "conglomerate discount" would be appropriate for investors trying to make sense of things.

Musk said Tuesday that combining the two companies he leads "would be a good idea," following months of speculation about the move. Both companies had formed special committees to discuss it, he added.

The companies do have business links. Tesla is selling cars, batteries and power infrastructure to SpaceX, and they are working together on Musk's Terafab, a chip-making venture that could benefit both. Tesla's recently unveiled flying Roadster uses SpaceX thrusters. Tesla has also invested $2 billion in SpaceX.

From a financial and investment perspective, though, it is hard to see why the two companies need to be more than just each others' partners and customers.

Take Tesla's Roadster. It isn't expected to start production until 2028 and won't be produced in high volumes. More important, the SpaceX thrusters themselves could be little more than a gimmick with limited real-world use.

Other potential synergies lie well into the future. Tesla's humanoid robots, for instance, could have a role to play in space ventures. AI technology developed at SpaceX could have applications across the company. Solar and battery tech would likely be helpful on Mars. Good luck factoring any of this into an investible model, though.

Terafab might represent the clearest near-term opportunity for the whole company to benefit from joint investment. But it is no easy lift. Musk said Tuesday the venture aimed to make its first test chips next year, and the aspiration was to reach volume production in the latter part of 2028. But he assigned "big error bars" to that possibility, given the complexities of standing up new chip production. Considering Musk's track record with ambitious timetables, investors should take this one with a mountain of salt.

Big baskets of businesses can be investor-friendly if the components share certain financial characteristics. At Berkshire Hathaway, for example, stable businesses with reliable cash flows predominate.

But the investment case is more complicated when the dynamics are vastly different between units, as is the case with SpaceX's stable of mostly tangentially related businesses.

In a sum-of-the-parts analysis of SpaceX in August, Deutsche Bank analysts assigned a base case enterprise-value multiple of five times revenue for the company's AI business. But the analysts gave an enterprise-value multiple of 58 times revenue to its connectivity business, which contains its Starlink satellite-internet service.

That makes the whole harder to value. Should investors give more weight to the capital-intensive AI business, or the steadily growing internet-service arm? How should they factor in the social-media platform X, or the AI-model development efforts?

Deutsche Bank's analysts are still bullish. Based on their sum-of-the-parts valuation, they put a price target on SpaceX of $235 a share -- about 47.5% higher than where the shares traded Wednesday.

But another way of reading this analysis is that there is a big conglomerate discount already present in the shares, and adding Tesla could make that worse. This would suggest that a merged company would likely be worth less than their combined current value of about $3.66 trillion.

Tesla shareholders have the most to lose in that scenario. A merged Tesla could be called upon to support other parts of the conglomerate if they struggle. That risks draining it of resources that would otherwise have fueled its growth.

SpaceX is already a conglomerate, having taken over Musk's AI operations and social-media platform earlier this year. Tesla remains more focused -- at least by comparison -- with electric vehicles, solar power and electrical infrastructure in its portfolio. Those pursuits have clear overlap. But already, the bull case for Tesla is that its bets on robotics and self-driving tech are the real drivers of its future value. With Musk, things just keep getting more complicated and speculative.

At Berkshire Hathaway, Wall Street's skepticism of sprawling conglomerates was overcome by overwhelming confidence in Warren Buffett himself. Something similar is clearly at work in Musk's strong following.

From his perspective, the merger would roll up his business interests into a tidy ball -- something he has been pursuing for decades.

As far back as the dot-com boom, Musk envisioned his financial startup X.com as a one-stop shop for transactions and investments. He revived the idea after buying Twitter and renaming it X in 2022, positioning it as a super-app that could handle communications and financial transactions.

Yet today X remains a place to post and little else. Musk's ambitions, always grand, don't always translate into reality. A merged, Byzantine conglomerate of his ventures would be the biggest bet yet on Musk, his vision and his management. Many investors will thrill at the opportunity, but many others will understandably recoil.