A Financial Times article recently described the structure of the SoftBank Vision Fund.

The details reveal breathtaking audacity in terms of SoftBank laying claim to investor returns without transfer of an equivalent level of risk. To avoid competing accounts, I will use the FT as my source of information, as I am not so much interested in what SoftBank is doing as I am in how investors might analyse such structures.

The tech and innovation focused fund has gained fame due to its size, currently a reported US$93 billion in commitments. Less broadcast is that SoftBank’s 44 per cent internal rate of return over the past 18 years is driven predominantly by two investments, Alibaba and Yahoo Japan. But the issue is not about investment ability, it is about whether the structure is fair.

Now to the main issue with regards to the structure of the fund – investors not only invest equity but must also provide debt. Although private equity (PE) funds often do leverage their investments, they do so by approaching lenders separately, not asking investors to provide the leverage.

The investor-provided debt is through a special class of units in the fund called preferred units. The normal investments are equity units. For each $100 an investor invests in the fund, $62 will be in preferred units and $38 will be in equity units. The equity units are the normal investment in a PE fund, whereby the investor gets a return made from buying and selling companies minus the fees paid to the manager. The preferred units are obliged to pay 7 per cent per annum and return the principal, regardless of the performance of the underlying investments.

On the face of it this looks like a good structure for investors, as part of their investment mirrors the performance of the underlying portfolio and part is safer, giving a 7 per cent coupon over the life of the fund and then returning the principal regardless of how the underlying portfolio performs. There are several problems with this type of thinking.

First and foremost, who is guaranteeing the coupon and principal on the preferred? The idea is that if the portfolio performs badly then investors will get a coupon and their principal on the preferred. But from the announcement nobody is guaranteeing that. That means that if the fund loses all of its money, then the preferred units are worthless. This is a serious risk as we shall see.

So in an extreme negative situation, the preferred units and the equity units have the same risk. Not good. What about a negative situation but not that extreme? Let us say that the investments return 0 per cent. The fund won’t have to pay a performance fee but will have to pay the 0.7 per cent to 1.3 per cent per annum management fee as well as 7 per cent on the preferred. The following maths assumes that total commitments reach $100bn and that, as announced, SoftBank’s $28bn investment is equity only. Therefore, $44bn will be in preference units. The life of the fund is 12 years and I assume a full life cycle.

In this case the preferred units that need to be paid amount to about $37bn over the 12 years. The management fee on the equity units would be $6.72bn. Some people will argue that investments don’t happen immediately and can exit early. I argue that there are plenty of fees that get taken out of each transaction, such as broker fees, as well as deals that don’t happen but money is spent on due diligence, such as legal and audit. In the end the ball park figure is scary and here it comes.

The coupon and the fees come out to $43.72bn. Since the medium-case scenario has the portfolio with a 0 per cent return this means of the original $100bn only $56.32bn is left. Of that $56.32bn, $44bn has priority payback for the preferred units. This leaves $12.32bn to pay back the original $56bn of equity units, or a total loss of 88 per cent even though the actual investments were break-even. How scary is that?

One can always crunch the numbers in a myriad ways and make things look good. But it is hard to argue that a structure that gives investors a loss of 88 per cent on their equity investment when the underlying portfolio is break-even is a good thing. Not only is that a terrible structure, to even try to argue it isn’t crosses the line.

This article was originally published in The National.