The Japanese conglomerate SoftBank shook the tech and investment landscape over the last two years with its $100 billion Vision Fund. Previously, a good-sized venture fund was a fraction of that amount — maybe $200 million tops.

This week, SoftBank announced Vision Fund 2, which aims to raise $108 billion. This fund could have a huge impact on the direction of future technology. It’s focused on artificial intelligence and data analysis. And of course, it’s making waves in venture capital all over again.

Host Molly Wood spoke with Paul Kedrosky, an investor with SK Ventures. She asked him how big the SoftBank effect has been. The following is an edited transcript of their conversation.

Paul Kedrosky: It’s massive. Because of the effect of having giant capital, a Death Star of capital out there, that could take companies who were primed to go public — let’s use Uber as an example — and inject enough capital into them that you could absorb the losses that a company like Uber was generating pre-IPO, and continues to generate post-IPO for what it’s worth. So, what this allowed companies to do was they were already priced, in a sense, as public companies, despite being private. But they hadn’t learned the capital discipline yet, so continued to generate huge losses. Once they go public, the stock doesn’t do much — and as we’ve seen in the most recent headlines from Uber, might end up doing some sizable layoffs.

Molly Wood: Uber is probably the best-known example. But you must assume that if the minimum investment amount from a SoftBank fund was $100 million that this fund may have also — and may continue to — be propping up some other businesses that might otherwise fail fast.

Kedrosky: Absolutely. The problem you run into is you become motivated by check size, which is a weird way to do venture capital. Specifically, that means I don’t say, “Is this a great business?” I say, “Is this a business to which I can write a check large enough for me that I can move the needle and allocate my money fast?” Because you must think: “I can’t write checks $3 million or $2 million at a time if I’m running $100 billion fund. I’m just going to end up with way too many portfolio companies.” So, in a weird way, you need companies that need a lot of money. And then I can just happily push them along down the road, and they end up being these sorts of perpetually money-losing companies that never go very far — in part because SoftBank needs them that way.

Wood: It’s like the Brewster’s Millions approach to venture capital. But you start to wonder: What happens to the landscape when a smaller company might not get funded purely because of this distorting effect?

Kedrosky: On that, I’m somewhat optimistic. It leaves an awful lot of ground open for relatively smaller funds that are running less than $100 million, because the companies that they’re likely interested in are just completely untouchable to funds at this size. Whether these funds can make any money doing that thing is a completely different question. But at least you’re not competing with SoftBank doing that. The people who are in the most trouble are the people who are raising these so-called mezzanine funds. Kleiner Perkins recently hived off a group that’s now doing later-stage venture. And those are the people that are up against SoftBank.

“What we don’t know isn’t very specific. We don’t know what Microsoft’s and Apple’s commitment is. What we do know is that some of the sovereign wealth funds have disappeared and new smaller ones have popped up instead…”

-Paul Kedrosky

Wood: When you think about the impact this has on the companies that are going to be part of our daily lives, the idea they could exist regardless of whether they ever make money [is] a little dark.

Kedrosky: It is a little dark. The silver lining in all of this is to look back at the amount of money that was squandered and blown up in in the late 1990s on things like dark fiber and stringing fiber off the cable all around the planet. That was all done without regard for economic return. And now we’re huge beneficiaries of that, because it turns out that lighting all that fiber has been able to drive down the cost of you being able to stream video, Netflix and other services. But you could make that argument that maybe something similar is happening today. That doesn’t sort of mitigate the problem that a bunch of people are going to lose a lot of money on the way.

Wood: What do we know about where the money is coming from?

Kedrosky: It’s a funny structure, and not necessarily funny “ha ha.” It’s funny in the sense that something like a third — and obviously SoftBank can change this on the fly and make everything I say here wrong — but roughly it looks like about a third of the fund is coming from SoftBank itself, which as a starting point is pretty unorthodox. So, you imagine a typical venture fund of $300 million that then told you that $90 million, pushing a third of that money, was coming from themselves, you’d say, “Well that’s interesting. You must be really committed to your own strategy, must really like the kinds of things you’re doing, because you’re investing your own money in it.” Well, yeah, but in this case, this 30 percent is coming from a mix of previous portfolio firms listing that may list publicly and may generate money and may in turn have some of those proceeds put back into the new fund. So, it’s a lot of provisos and what-ifs and wherefores attached to this one-third. It’s not as if the partners in SoftBank Vision Fund itself are putting the money right back in.

Wood: OK, you’re right — that’s funny-weird. Beyond that, we just don’t know very much about it. We know that Microsoft is involved in some way, some of the previous partners like Apple and Foxconn are involved. But we don’t know much more than that.

Kedrosky: We don’t know to what scale. So — I’m just doing the math here: Take $100 billion, subtract $30 billion, you got yourself $70 billion left. So, we got Microsoft, we got some large Japanese banks and some other investors like Apple and Hon Hai, which is Foxconn, an Apple assembler, who’ve all made sort of memoranda of understanding saying that they will commit to this fund. We have no idea to what size or scale. Perhaps more interesting/troubling is some of the largest investors in the prior fund, which were Middle Eastern sovereign wealth funds, haven’t made any commitment to the new fund. And instead we have, I think it’s Kazakhstan’s sovereign wealth fund that’s making a small commitment. So, it’s really the nature of the investors. What we don’t know isn’t very specific. We don’t know what Microsoft’s and Apple’s commitment is. What we do know is that some of the sovereign wealth funds have disappeared, and new smaller ones have popped up instead who will likely only be able to make very small investments. So, this is a very, very squishy number where you end up.

Wood: I feel like you may be saying that this fund is not a done deal. Should we not be counting it as money?

Kedrosky: I would agree with me on that. I would say this is closer to a marketing document than a funding document — by which I mean they’re saying: “Here’s what we’d like to do. I hope we can get a lot of people excited about it. We don’t have all the money raised. Hey, are you interested in putting up some money? And by the way we need a lot, because we’d like it to get to $100 billion.” So, in that sense this is much more about marketing than anything else.

Wood: So, then what is the arithmetic of a fund this size? Is there any possibility that they ever make a return that’s bigger than the investments?

Kedrosky: It’s entirely dependent on the public market. This is the issue. Let’s say it’s a $100 billion fund. They’re going to have to return a reasonable number to their investors — let’s say in excess of 18 to 20 percent, which is roughly what they should be hoping for in terms of an annualized return to be competitive with public markets, and there’s a thing called an illiquidity premium and all this stuff. They’re going to have to generate roughly for X the initial capital. So, call that $400 billion.

And the way the venture industry works is that most of your returns come from less than 20 percent of the companies in your portfolio, so, in this case, say $20 billion of what they’ve allocated or relatively a much smaller number of companies. So that means you’re going to have to see some series of Uber-sized IPOs out there that generate a huge amount of proceeds for the proceeds for SoftBank. The math never works to generate a reasonable return, to come up with this $400 billion of returns that you need to provide investors with, a market-competitive return on a fund this size. That may turn out to not matter to people because there’s some perverse phenomena in the sovereign wealth fund business. They just want to write large checks. They had the similar problem. So, Kazakhstan, let’s take it as the example sovereign wealth fund. They have their own fund, and they have a target allocation — it’s called alternative investments — which includes venture capital. So, if someone is willing to let them write a large check, that solves a huge problem for them, because they got a lot of money to invest. The problem that they might not actually generate returns turns out to be secondary to the problem being able to write large checks. This is a funky, funky business out there.

Related links: more insight from Molly Wood

You heard Paul Kedrosky say that SoftBank will invest some of its own money and proceeds from the Vision Fund into this second big venture fund. Well, some of that money could come from the Sprint and T-Mobile merger because — fun fact — SoftBank Group, the parent company of all this, owns Sprint.

The Nikkei Asian Review has a story about how the merger, once it gets final sign-off, will take a big albatross off SoftBank’s neck.

SoftBank Group bought Sprint in 2013 for $20 billion, always planning to merge it with T-Mobile. But obviously, that’s been a long road and Sprint has been losing money all along. If the merger does finally happen, the paper says it’ll free up a bunch of money for SoftBank to reinvest in Vision Fund 2.

A site called Market Realist has an interesting profile on Masayoshi Son, the founder of SoftBank and the guy who personally makes a lot of these investment decisions. It argues he’s playing the long game and has actually created a 300-year business plan. Son and SoftBank lost $70 billion in the dot-com bust of the early 2000s … which is apparently just play money at this stage in the game.