long but fun read

–

A company that wants to borrow money has a few options. It can borrow from a bank. Or it can borrow from non-bank investors. Or it can issue bonds and borrow from the market.

Borrowing from a bank is a bit uncool. Banks are risk-averse, might need more security than the company might want to give, or just unwilling to lend to it.

So the company goes to an investment bank! If the company wants to borrow money quickly and is cool with paying a fair bit of interest, the investment bank might find a willing private lender. But if the company wants the best, market-decided interest rate? No problem, the investment bank can then help the company do a debt offering. It will help price its bonds, market them to investors, and make sure the debt offering goes through.

The investment bank is the middleman. It takes no risk, because it does not lend to its client. [1] It just ensures that someone out there is willing to take on that risk.

Anyway, here’s what I wrote a few months back:

Investment banking is notoriously competitive. You’re constantly competing with tens of other banks looking to do deals. In most cases there is little to differentiate you from the bank across the street. You probably hire people from the same set of colleges, suck their souls the same way, and the suckers then move around within the same set of banks.

So if a client comes to you and tells you that he wants to do a plain old vanilla debt offering—what exactly is the differentiator that you’re going to offer? The client company wants to sell bonds to investors, give them a fixed coupon rate, and it wants a bank to prepare the documents, do the disclosures, market the bonds to investors, etc. It doesn’t get simpler than that!

An investment bank’s main job is relationship-building. And part of that relationship-building is going overboard with what it promises its client. Last month, SEBI issued an enforcement order against Axis Capital, [2] yet another investment bank which became overzealous in its attempt at winning business. Here’s what happened:

1. Sojo Infotel, a small no-name tech consultancy, wanted to borrow money by selling bonds.
2. It didn’t want to use that money to run its business! It wanted to buy shares in another company. Presumably, the interest that it would have to pay to borrow for something as risky as this would be high.
3. But the interest rate wasn’t high! Axis Capital, Sojo’s investment bank, guaranteed to repay the bond investors in case Sojo defaulted. Because of this guarantee, Sojo had to pay a coupon rate of 8.48%—that’s only marginally higher than a fixed deposit these days.

Axis Capital was the middleman helping Sojo raise money from bond investors. Axis Capital was also a participant guaranteeing the same bonds. SEBI doesn’t like that.

Boring Money is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.

The mechanics

Sojo borrowed ₹260 crore ($31 million) by selling bonds. It then used part of that money to buy shares of Lava International, a mobile phone company. The funny bit though is that the people that owned Sojo, also owned Lava. Of the ₹260 crore, the first ₹13 crore went to buy out another investor in Lava. ₹100 crore went to Lava itself. Eventually, Sojo ended up owning 2.5% of Lava for ₹113 crore. Not sure what happened to the rest of the money.

The plan was that Lava would go public in a definite timeline—15 months. Sojo could then sell its shares and use that money to repay its bond investors. Leaving aside Axis Capital’s guarantee, whether Sojo’s bond investors were paid or not depended on Lava going public.

But these were secured bonds. If Lava failed to go public or if Sojo failed to repay for whatever reason, the investors needed something to ensure that there was still a way they could recover their money. Sure, the Axis Capital guarantee was that way, but that was the weird stuff that the investment bank did. On paper, the bond investors needed other collateral.

That security was… Lava shares! If Lava failed to go public, Sojo’s bond investors would end up getting Lava shares! [3] And those shares, considering the company would have just failed to go public, couldn’t be worth a lot. Of course, none of this mattered. No matter how dumb the security backing the bonds was, Sojo still sold the bonds at a coupon rate only a tinge higher than a fixed deposit.

Because Axis Capital guaranteed Sojo’s bonds!

Playing both sides

Another oddity is how the agreement was structured.

Axis Capital, the investment bank, represented Sojo. Another Axis company, Axis Trustee Services, represented the bond investors. This is normal stuff. A group of investors needs a trustee to represent its interest and hey Axis was already involved so no reason for it not to be Axis Trustee. Well, here’s a bit from the bond offering document:

Notwithstanding anything to the contrary contained in any of the Transaction Documents, upon, occurrence of an Event of Default, an Asset Sale Event shall be deemed to have occurred in respect of the Debentures. Upon the occurrence of an Asset Sale Event, the Parties irrevocably agree that that the Debenture Trustee shall be bound on the instructions of the Majority Debenture Instructing Group to intimate the Issuer of the occurrence of such Asset Sale Event and Axis Capital shall (as Sale Arranger) be entitled to sell any of the Secured Assets…

If Sojo defaulted, or if Lava didn’t go public in time, two things would happen:

1. Axis Trustee would intimate the issuer of the bonds, that is Sojo.
2. Axis Capital would be entitled to sell any of the secured assets. That is, Lava’s shares.

Oh, Lava tried but failed to go public, so all of this actually happened. Here’s SEBI describing the events:

On October 27, 2023, Axis DT invoked pledge over 26 % shares of LIL held by its Promoters, which were pledged by the promoters of Sojo as security cover for NCDs issued by Sojo. As ACL was unable to find a purchaser for the pledged shares, it led to triggering of Asset Purchase Event on March 15, 2024, which required ACL to fulfil its “underwriting commitment”

Axis Trustee, since it represented investors, got hold of Lava’s shares since they were the collateral against the bonds. But then, somehow, Axis Capital had to find a buyer for those shares? Axis Capital was the investment bank representing Sojo! It wasn’t working for the bond investors or for the trustee. It seems to have just turned around, unmasked itself, and suddenly started working for Axis Trustee instead? [4] [5]

Hidden debt

Axis Capital took a shitty deal it wasn’t legally allowed to. Sure, one reason it did that was to win Sojo’s business. But that wasn’t all! The larger reason seems to be that it wanted Lava’s business.

Lava was to go public within 15 months of the Sojo bond offering. A company that wants to go public needs an investment bank. If Lava’s owners had already used Axis Capital to borrow money at an unusually low interest rate, of course they would then use Axis Capital again to take Lava public. (Axis Capital was, in fact, the lead manager for Lava’s failed IPO.)

I wouldn’t be surprised if this entire episode happened the other way round. Axis Capital was to be the lead manager for Lava’s IPO, but Lava needed money. It could be bad for the IPO should it borrow directly—debt looks bad on a balance sheet—so Axis Capital figured that Lava/Sojo’s owners could just use Sojo as a proxy to take on debt and pass on the money to Lava.

For now, SEBI has restricted Axis Capital from taking on any more debt related assignments. There might be fines later, but this is all just the securities related stuff. But the meat is in the banking related stuff. That’s something SEBI has left for the RBI.

Axis Bank [6] owns Axis Capital. If Axis Capital guarantees debt in a distorted deal such as this, it is effectively extra risk that isn’t showing up on Axis Bank’s books! Axis Capital just lost ₹174 crore because of Sojo, and SEBI has identified at least 5 other instances where it did the exact same thing. If it happens to lose similar amounts on all of them, it could be more than ₹1000 crore ($120 million) of lost money.

RBI should be all over this very soon.

–

Footnotes:

[1] Okay, not no risk. An investment bank underwriting a bond offering might take market risk. If the bond offering does not get enough subscribers, it can buy some of the bonds so that it can sell them later in the open market. But this risk is part of the service!

[2] SEBI’s order was inspired by a blog post! Is Axis Capital an Investment Bank or a Hedge Fund?

[3] Lava shares wasn’t the only security, but it was the main one. Other security included Sojo’s shares, Sojo’s assets, and also a personal guarantee by Lava/Sojo’s owners.

[4] Another matter of concern here is… why was Axis Trustee okay with outsourcing its responsibility to Axis Capital? Would it have done this had the two not been part of the same Group?

[5] A slightly facetious way to look at this is, maybe Axis Trustee just wanted to save money? Axis Capital trying to sell Lava’s shares was a farce anyway. Someone buying the shares of a private company does so with the hope that the company would go public in the future and they would make money. Who would want to buy the shares of a company that provably would not go public? Since Axis Capital had to pay up anyway, Axis Trustee might as well save some cash by asking it to do the formality.

[6] Disclaimer: I own a teeny amount of Axis Bank stock. I should probably sell it off so that I don’t need to include this dumb disclaimer again.