If NVIDIA were to issue a convertible bond...
by Laurent Le Grin 23/8/2023

This is a fictional dialogue between the fictional CEO, IR and CFO of NVIDIA. Nothing that you will read next constitutes an investment recommendation.

Minecraft & Bitcoin

CEO (Lisa): Good morning, everyone. As you're all aware, our shares have experienced a strong rally recently, and we've been exploring various ways to capitalize on this positive momentum. One option that has come up is the issuance of a convertible bond. This could potentially allow us to secure additional funds while taking advantage of our elevated stock price. John, I'd love to hear your thoughts as our CFO on this matter.

CFO (John): Good morning, Lisa, and everyone. Indeed, our stock has been performing exceptionally well, and it's a great opportunity for us to consider our next financial move strategically. Issuing a convertible bond at this juncture could be a smart move, especially if we can structure it in a way that aligns with our long-term goals. The key question is how we can effectively "game" the investors' perception to get the most benefit for our shares.

CEO (Lisa): Thanks for your honesty, John. Let's dive into the specifics. How do you envision us structuring this convertible bond to maximize our benefits? We want to ensure t hat we're not only getting the necessary funding but also positioning ourselves for future growth.

CFO (John): Absolutely, Lisa. I'm thinking we should focus on two key aspects: the conversion premium and the conversion ratio. By setting a conversion premium that is attractive to investors but not overly dilutive, we can entice them to convert the bonds into shares at a price higher than the current market value. This way, even if our stock price dips slightly after conversion, we're still in a favorable position. As for the conversion ratio, we should carefully consider how many shares each bond can be converted into. A slightly higher ratio could encourage more investors to convert, potentially boosting demand for our stock.

CEO (Lisa): Those are solid points, John. We want to strike the right balance between incentivizing conversion and maintaining shareholder value. I'm also thinking about the messaging around this move. How can we present the convertible bond issuance to investors in a way that makes it seem like a strategic move rather than a reaction to our stock's recent rally?

CFO (John): That's a crucial consideration, Lisa. We should emphasize the purpose behind the convertible bond issuance, highlighting how the funds raised will be used to fuel our ongoing innovation and expansion initiatives. We can also point out that our stock's rally reflects investor confidence in our growth prospects, making this an opportune time to secure additional capital. By framing it as part of our broader strategic plan, we're more likely to gain support and enthusiasm from investors.

CEO (Lisa): I like that approach, John. It's essential that our communication remains transparent and aligns with our company's values and long-term vision. Before we proceed, let's also engage our legal and investor relations teams to ensure we're compliant with all regulations and that our messaging resonates well with both our existing shareholders and potential bond investors.

CFO (John): Agreed, Lisa. Legal and investor relations' insights will be invaluable in navigating this process smoothly. With the right strategy in place, I believe we can issue a convertible bond that not only capitalizes on our stock's rally but also positions us favorably for the future.

CEO (Lisa): Thank you, John. Let's move forward with developing this strategy collaboratively, incorporating the input from all relevant teams. I appreciate your expertise and dedication to ensuring our company's financial success.

After a few hours of dedicated work, the NVIDIA teams come back with a few ideas.

Investor Relations (Gwenn): A frequent concern for companies is how to classify debt and equity. Convertible bonds, which have features of both, are especially confusing because of differences in accounting for seemingly similar instruments. In response to this, on August 5, 2020, FASB released an update under ASU 2020-06. The update aims to make it easier to understand and implement the current guidance on financial instruments that have liability and equity characteristics. Under this model, the premiums are recorded as paid-in capital. Under ASU 2020-06, for a convertible debt instrument issued with a substantial premium, these premiums will continue to be separated from the debt instrument and recorded as an equity component. Some may consider a premium equal to or greater than 10% of the par value of the host debt instrument to be substantial. However, a 10% premium is not a bright line; all relevant facts and circumstances should be considered to determine whether the premium is substantial.
To address any concerns regarding dilution, I suggest considering a premium of 30% or more. Given the strong potential of our company through crypto mining, video games, and now AI, it's evident that even with valuation levels of 40x 2024 sales or 26x 2025 sales, investors are highly enthusiastic about us. John, Lisa, let's dive into some calculations and explore the feasibility of convertible bonds. In my opinion, we could bolster our available cash by issuing a substantial $2 billion worth of convertible bonds to facilitate opportunistic share buybacks. While awaiting the right opportunity, we can still invest the cash in short-term options yielding 5%.

CFO (John): Go ahead, I'm sure this is going to be informative.

Investor Relations (Gwenn): OK. First of all, the initial aspect I verified is the process the bankers will use. They'll gauge interest by reaching out to significant hedge funds and a few long-only convertible players over a day, and in the evening, they'll come back to us with investor interest. I propose going with $2 billion US dollars for 7 years. I suggest a coupon rate of 0.5%. In our discussion, we'll explore how we can limit the dilution risk by offering a reasonable 30% premium, so as not to spook the market with 50%. But I'll present a way to synthetically increase the conversion premium.
Regarding our valuation, our company is at the top of the tree, and we're seeing the monkeys' behinds. However, it's likely that bond investors will jump at the opportunity, especially given our transparent and promising business: AI, video games, robotics, deep learning, crypto mining. We have quite a story to tell.
Additionally, NVIDIA doesn't pose many ESG risks, even though there are numerous negative externalities (intensive server usage, health risks of video games, morality of cryptocurrencies, etc.). Managers are primarily focused on carbon footprints. Anything related to the morality of cryptocurrencies and health impacts from games don't really stand out as significant risks. I know that a friend's son is being treated for severe depression. He had a breakdown after his Minecraft account was banned for the third time. The child couldn't handle it. Video games are probably a significant future health risk. I even wonder if the Chinese are right in limiting screen time. However, by then, investors will likely see us as an AI company.
I do want to mention the SEC's reprimand that NVIDIA failed to disclose in 2018 that a significant portion of its revenue came from cryptocurrencies rather than directly from video games. We've reached an agreement with the SEC to pay $5.5 million to settle the matter. The dispute is well documented in the 2023 annual report (p.68 2023-Annual-Report-1.pdf (q4cdn.com)). No provisions have been set aside for litigations. We might have questions on this topic. But as for other aspects, such as user health, it's unlikely.
Considering recent issuances in the US convertible market, I believe, as I mentioned, that we could aim for a coupon rate of 0.5%, maybe even lower, for a maturity in 2030. I have a friend at AKAMAI who issued a convertible bond for 2029 with a coupon rate of 1.125%. I haven't contacted him to avoid giving away our intentions. If AKAMAI is paying 1.125%, at the very least, we shouldn't have to pay anything. For your information, if we were to issue a vanilla bond, we'd probably have to pay something around 5%.

CEO (Lisa): Okay Gwenn, very clear. Can you show us what this instrument would look like in your opinion?

Investor Relations (Gwenn) : Yes, here is the animal we're talking about.

Coupon Rate 0,50%
Maturity 7 years
Effective yield 5%
Premium 30%
Stock Price $ 500,00
Par Value $ 1 000 000,00
Proceeds : $2 000 000 000,00
Conversion Ratio: 1 538,46 shares
Parity (%) 76,9 in percentage of par
Bond Floor (%) 74,0 in percentage of par
Use of proceeds : Buy-Back, call spread

My grand idea is that we could use the raised cash for share buybacks, but not at $500. In our discussions, we all agree that nobody understands what Powell is saying, our country's debt has reached the top of the tree with the monkeys, the war in Ukraine drags on, we'll likely face regulation on AI head-on. China is also slowing down significantly. In short, there's a heap of crap piling up on the horizon. So, we issue our convertible, invest the cash at 5% for 3 months, and we might re-invest it at an even higher rate. The Fed is chasing the curve like my Golden Retriever chases after a breadcrumb. They're capable of wrecking everything.
To limit the dilution risk, we can buy a call spread to increase the premium from 30% to 50%. John and Lisa, you know that volatility is hard to predict. The legs of the call spread are valued using Black & Scholes, I have used a calculator available on the web, considering an average volatility of 55% (the average of ESPP volatilities) and a maturity date of September 30, 2030. In any case, our investors are willing to pay us 40 times our sales, so a volatility of 55% is not an unreasonable measure.

Call 130%: $275
Call 150%: $260

For one share, the cost of the overlay is therefore +$225 - $240 = -$15.

The anti-dilution hedge won't cost us much.

As the premium can be considered substantial, the instrument is accounted for in two components:

one equity component representing the conversion option and a bond component representing the present value of the coupon payments (2%) discounted at 5%, which is the market interest rate.

Equity portion = Total Proceeds - Present Value (Coupon + Redemption Value) = $2,000,000 - $1,479,226,394 = $520,773,606.

Therefore, the cash on the balance sheet increases by $2 billion - $46,153,846 (cost of the call-spread overlay), resulting in an increase of $1,953,846,154 due to the successful issuance of the new convertible bond.

This money won't cost us anything since we'll invest it at 5% while proceeding with the share buyback. The bond leg is the sum of the coupons and the nominal value discounted at 5%. The portion related to the conversion option is therefore equal to: $2 billion - $1.479 billion = $521 million.

As time passes, the interest rate differential between the coupons and the effective yield will gradually increase the book value of the bond up to its redemption value, in other words, 100% of par value.

The conversion option, on the other hand, will be re-evaluated at each accounting period. If the value of NVIDIA's shares continues to rise, the equity will increase along with the increase in the convertible bond's conversion ratio. From this, we can draw an interesting conclusion: the company deleverages as the underlying value of the convertible bond increases.

Now that we've seen how the convertible bond will be accounted for and its impact on the balance sheet, let's discuss how the share buyback is accounted for.

The share buyback is not a complicated operation, unlike the recording of a convertible bond. Assuming management is authorized to do so, cash will decrease based on the pace of buybacks in the market. The reduction in cash will be offset on the balance sheet by a decrease in the number of outstanding common shares and distributable earnings. The repurchased shares themselves do not disappear but are held in reserve as part of the capital.

CEO (Lisa): Thank you, Gwenn. You've used $500 as the price, I see. I'm not sure if our results tonight will allow us to achieve that. But I'm inclined to closely monitor the situation after our results. So, Gwenn, if I understand correctly, we could repurchase $2 billion worth of shares at $400 or $300, I might be off here, and we would gain $600 million or $300 million in equity if the market takes a downturn for some reason?

Investor Relations (Gwenn): Yes, that's right. Without paying anything, because even if our stock shoots up, we'll have had the opportunity to buy the shares and thus avoid dilution.

CEO (Lisa): John, we're talking to the compliance officer and legal this afternoon, and let's meet for the post-results call this evening.