Amazon Is Seeking $16B Bond Sale for Whole Foods | Hacker News Reader

Amazon Is Seeking $16B Bond Sale for Whole Foods

bloomberg.com

79 pointsmcone57 comments

I'm still surprised more companies aren't doing it. Investment grade companies can get 0% in the Eurozone, and eurozone junk bonds have lower yields than US treasuries. Because they're safer? Who cares just sell the bonds.

Even at 0% it's not risk free on the part of the company. Unless you have EUR cash inflows, you'll be shouldering FX risk to convert to EUR (say from USD) to make the loan payments.

This past year alone has seen a range of about 13% for USD/EUR. A shift like that would wipe out any interest related savings.

FX risk is extraordinarily cheap to hedge against, the futures market is really deep.

The cost of hedging forex is at least equal to the interest rate difference between the currencies.

Can someone who knows how that works do the math for us?

It's an arbitrage issue. If you can borrow EUR for 1% and it costs %1 to get a USD/EUR hedge, you can turn around and loan USD for 2% and be guaranteed to not lose money. The market eliminates arbitrage opportunities whenever they appear; so if the cost of borrowing EUR is less than the cost of borrowing USD, you'll have to pay at least the difference in order to get someone to guarantee a rate for converting USD back to EUR later.

There could in theory be other risks in EUR which prevent the arbitrage (legal? liquidity? cost of initial transaction being high?) which is why I'd prefer to see actual numbers bearing that out instead of just a derivation from first principles.

In general the price of a futures contract is that which avoids arbitrage opportunities.

If the interest rate in EUR is 2% and the interest rate in USD is 4% and the cost to maintain a EUR/USD hedge is less than 2% then you can make free risk-free money by borrowing EUR at 2%, exchanging it to USD, and lending it out at 4%.

Such a situation is unstable as more and more capital would pour in to take advantage of it and the cost to hedge would rise or the interest rates would move closer together until the gap disappeared.

Collars in the options market

Futures

or plain old fx swaps solve all of these problems

How much would it cost to hedge the EUR equivalent of USD 16bn of these bonds?

Its different for every product so lets talk about the futures market as it is the most transparent and [likely] deepest:

According to the US government there is currently 57 billion EUR in open futures contracts, which is currently 68 billion USD. A new market participant that needs 23% of this current level of liquidity should be able to attract the interest they need without moving the market.

http://www.cftc.gov/dea/futures/financial_lf.htm

Qualified participants are able to negotiate non-standard sized futures contracts, and wouldn't have to worry about broker commissions.

In the futures, the rolling transactions would have to be maintained.

IDK technically it should be near free, but the other banks are pretty greedy so hard to say. Really shouldn't cut into the free-ness of the yield and bond issuance.

I don't know how you can suggest issuing in EUR if you cannot compare the USD interest expense vs what you think they issue under EUR.

I'm not trying to be rude but I don't think you know what you are talking about if you are suggesting that executing a futures transaction that large won't move the market and that rolling USD 16bn equivalent futures is going to somehow come out near free. Amazon merely announcing an EUR deal that large is going to move the market and that's not even considering actually executing these theoretical trades. You don't want to give people the ability to frontrun you on something that large.

Suggesting non-standard sized future contracts completely negates the liquidity argument and also leaves you at the mercy of either your counterparty or someone else who likely has the ability to move the market in order to screw you over.

I'm pretty sure Amazon along with other issuers who issued multi-billion USD deals this year took a look at issuing in EUR and decided it wasn't in their best interests.

> I don't know how you can suggest issuing in EUR if you cannot compare the USD interest expense vs what you think they issue under EUR.

Okay.

> ... liquidity

Liquidity begets liquidity. Do what you can.

> I'm pretty sure Amazon along with other issuers who issued multi-billion USD deals this year took a look at issuing in EUR and decided it wasn't in their best interests.

You shouldn't make that assumption. There are many market participants that wouldn't consider the world outside of their market. There was a long time where the US was the only liquid market, and still eclipses the single currency bloc for liquidity. When favorable inefficiencies come up it is easy to miss them.

The cost of the transaction should be a factor, but the foreseeable costs are not that big of a factor. The company can issue 10 - 40 year bonds (or whatever else the market can bear) at low, zero, or negative interest rates and not have to worry about it until 10 - 40 years later. And then what? Well they just issue more bonds, in the next low interest rate environment and use a lot of the proceeds to pay off the remaining bit.

A combination of this and EUR credit is just not as liquid as USD.

It looks like the deal is 16bn across 7 tranches. EUR equivalent would be much harder for the market to digest and a lot more difficult to hedge.

>I'm still surprised more companies aren't doing it.

Microsoft issued bonds[1] to fund the acquisition of LinkedIn even though MS had plenty of cash in the bank to buy it outright. (Not sure how much of $100+ billion is in USA vs overseas which may have made the $26 billion cash payment for LinkedIn a stretch.) Alternatively, Microsoft could have paid via stock -- but they didn't. (Or maybe Reid Hoffman negotiations said "no" to MS stock.) In today's climate of low yields, healthy companies can buy companies without stock or cash holdings. Just issue bonds.

Facebook bought Instagram and Whatsapp with stock so the calculus was different there.

[1] http://www.investopedia.com/news/microsoft-leads-aaa-bond-sa...

If your investments make a better return than you'd pay interest on debt, it makes sense to take out a loan. If I'm earning 7%, and the interest on the loan is only 3%, I should take out the loan.

Surely that's only if you can make more than the delta by investing the cash is something else? So in your example it'd make sense if you could invest in something that returns more than 4% (7 - 3)? Otherwise you could make more money by not taking the loan.

I am assuming you're not taking loans for things you don't want. Even if you're taking loans for operation expenses, it may make sense.

If you can get 90 days at no interest, for example, you might as well take it and get 90 days of growth on your investments.

It is about balancing possible risks as well. In this example you are able to get 4% in as close or near to risk free as possible.

Right, but the alternative for Amazon isn't liquid cash, it's Amazon stock, which they believe will increase in value faster than the interest on bonds (not unreasonable imo).

If they had $16B lying around in cash, yes that question would probably make sense. You also have to take into account where that money is, since you can't bring overseas money within the US without paying a tax.

Some great points, and I totally agree with all of it. The parent implied that it always makes sense in general to borrow if the return is greater than the interest:

> If your investments make a better return than you'd pay interest on debt, it makes sense to take out a loan

Which doesn't quite add up!

What doesn't add up? If your investments make more money than the interest you pay on debt, then conceivably you're making more money than you are paying in interest on the debt correct? The situation I imagine here is that if I have $X in cash and a house costs exactly $X, I would rather invest all of $X in a place that gives better returns than the interest on a loan for $X, because I will be making money.

> If your investments make more money than the interest you pay on debt, then conceivably you're making more money than you are paying in interest on the debt correct?

Yes, you're making money, but you're making less than you could have without the loan.

In the original example there was an investment that returned 7%, and a loan was available for 3%. You can choose to invest your own money (let's assume it's $100,000), or to take a loan. If you take the loan you're making $4000 a year (4%) from the investment, and if you didn't take the loan you'd be making $7000 a year (7%). You're worse off!

But you have still have your original $100k, so you can now go and invest that. Say you find another opportunity that returns 4%, and invest your money there. You're now making $8000 a year ($4k from the investment with the loan, and $4k from this new opportunity) You're better off!

But what if you can only find other opportunities that return 2%? If you were to invest your money there you'd be making $6000 a year, so you'd still be $1000 a year worse off for taking the loan!

My point is that it doesn't always make sense to take the loan. There's a trade off - you're taking worse returns in exchange for having additional capital available to you. If you can invest that capital in something that returns more than the delta between what the original returns would have been and the interest on the loan only then does it make sense. If you can't get better returns that the delta you'd be worse off - at least in terms of returns, perhaps that's something you're ok with to have additional capital available.

It also makes sense from a liquidity perspective. Amazon has 21bn~ cash on hand which could easily pay for the Whole Foods transaction. Even though Amazon has proven that they can turn on the FCF fountain when they want to, it still makes sense to hold onto cash.

Yeah, there's lots of places where it makes sense, even in your personal finances

When I sold my company, the financial agency with whom I worked gave me free services with a financial advisor - not the same as a financial manager. It took me a few minutes to get it, but you probably still want to finance a house - even if you can pay cash.

Your cash, when invested, can earn more than you're spending to service the loan, in some situations,

While there are individual EU HY bonds yielding lower than Treasuries, that is not the market as a whole. To be generous, 30 year Treasuries are currently yielding 2.84%. I'm looking at Bloomberg Barclays Pan-European High Yield index (ticker LP01TREU) and it currently has a YTW of 3.24% with an average life of 5.07. If you were to match that life to 5 year Treasuries, those are yielding 1.82% right now. That's not even touching on the currency and duration mismatches.

Yes, and Pan-European means there are 30 countries which can be isolated for inefficiencies. Outside of the European Union with its one central bank distorting things very publicly, there is the Swiss National Bank distorting similarly and steeper with much less publicity, there is an even steeper negative yield in Sweden, Denmark and more.

The "average" is a sideshow when there is so much opportunity if you know where to look.

> I'm still surprised more companies aren't doing it.

Almost every major company has been doing it for many years. Especially large companies with tons of cash overseas. Instead of repatriating cash from overseas and paying taxes, corporation borrow against their overseas cash and pay dividends to shareholders or invest.

Imagine if you were APPL and you have $1 billion a in tax haven in europe. You can bring it back to the US and pay 20% in taxes or borrow against that $1 billion and pay 1% in interest.

Edit: @seanmcdirmid

> Except it isn't so much the tax haven in Europe as it is profit stuck in china (their second largest market)

Their second largest market is Europe and has been forever except for a blip in 2015/2016.

And Apple's foreign headquarters is ireland.

Most of apple's overseas profits are registered and kept in europe.

"Apple’s expanding overseas cash pile has drawn scrutiny from the European Commission, which says the company should be paying corporation tax on profits from Europe, Africa, the Middle East and India in Ireland, where it registers those sales."

http://www.telegraph.co.uk/technology/2017/05/01/apples-cash...

Ireland and europe are tax havens for many of the silicon valley companies, including google, facebook, etc.

Except it isn't so much the tax haven in Europe as it is profit stuck in china (their second largest market). Same is true for Microsoft.

How can they borrow against that cash that's overseas? Wouldn't the lender be thinking that Apple has to pay 20% in taxes to repatriate the cash in the event of an emergency? Is Apple able to borrow more than 80% of the value of cash they have overseas?

Clearly more companies are using debt and using more of it. US corporate debt market is now around $8 Trillion dollars outstanding [1] and new debt issuance in 1Q 2017 was up 70% [2], and even this article mentions multiple examples of recent "mega bond deals" by other companies [3].

I'm not sure how Eurozone debt came up. I didn't read anything about Amazon and Eurozone issuance in the article, did I miss it? Though, technically Amazon did issue an unregistered private placement of debt in this case, so the debt was not eligible to be sold in the US or sold to US persons unless exempt of course [4].

As for the claim that investment grade corporate issues can get 0% in the EU or that junk bonds have lower yields than treasuries, or any sort of ideas like that where it is "cheaper" to borrow in Europe because spot rates are printing lower (and negative) there - that's a miscalculation and debt markets doesn't work like that. First, at a basic theoretical level there is a concept of interest rate parity which is maybe worth reading about but not really super important after freshman year of college [5].

I could not find a public non-paywall link with current EURUSD basis swap rates (basis swaps would help show the adjusted rates foreign borrowers pay) but the following links from the ECB show how much they are sucking right now suffering from low GDP and low inflation (remember debt that one borrows is paid back with future money that has inflated (and sometimes deflated) so if inflation is relatively high between the time money is borrowed and paid back that is good for the borrower) [6] [7]. According to this link option adjusted spreads on EU high yield are around 275 basis points not lower than treasuries, that is +275, as of right now [8].

While off the subject for a sec, it appears currency FX risk hedging also came up in this thread, with some people talking about hedging currency FX on a company's non-USD denominated debt being a good thing and/or inexpensive and/or easy for a company to do. IMHO that is a misleading interpretation of what corporate currency FX hedging is. A company "hedging" a financial risk like a currency depreciation going against them does not elimate risk nor lock in some sweet profits like people seem to think hedging accomplishes. What happens is future currency moves can lead to uncertain GAINS OR LOSSES for a company, but nobody knows which way it will go. So company management can choose to reduce their potential gains and potential losses from currency FX in the NEAR future in exchange for taking a certain definite loss right now. A company can only do this for the near future, 3-24 months or so, on a rolling basis if they choose. (Amazon issued for 40 year bonds, there is no such thing as 40 year currency hedging forwards, nothing is locked over this time, a lot can happen).

It's important to note a company will have paid money to eliminate downside and eliminate upside. The company has also added a new risk, the risk from the hedge project itself, as currency hedging doesn't always work as expected. Investment banks make mistakes (Citi and UBS are notably error prone, GS may or may not screw up on purpose) or central banks and gov regulators do weird things a FX hedge didn't factor in when started (Za Sviss). The list goes on.

I'm not anti corporate FX hedging, I see companies using currency hedging in the right way and to their advantage every day. I just was hoping to clarify that hedging is not an elimination of risk for a small cost. It's the reduction of potential profit and loss in exchange for a certain loss, this loss can be expensive and the process can go wrong making things worse.

Q: Is a certain relatively small cost every month better for a company than an uncertain gain or uncertain loss every month?

I would say sometimes. Depends on the company and the context.

One thing we do know is Amazon has foreign currency risk and Amazon has chosen NOT to hedge it or at least not hedge a huge chunk of it. Their latest SEC filing reports a loss of at least -$450 million due to foreign currency changes going against them.

Ok, sorry for the long post, bathroom break over, I just want to add one thing: companies cannot just borrow as much as they want for no reason just because rates are low. Many companies with low or no debt are not borrowing because they have no f'ing clue what to do with the money if they did take on debt. Companies need to do something with borrowed money AND they need to tell lenders more or less what that something is. (Amazon said clearly in their filing they want to borrow this money to buy a chain of overpriced health food stores, Apple has tons of cash saved up but Apple has made clear they are issuing debt anyway so Carl Icahn will shut up).

If a company can't think up a worthwhile making or selling of something new to invest their capital into (a problem many companies today are suffering from) company existing debt doesn't have to just sit in the bank earning nothing. The company can pay back loans early, or this money can be re-invested into other bonds! Yes, Debt on debt. Though these bonds a company's corporate cash invests in are going to be better-rated more-boring bonds lower yielding than their own (often just treasuries and short term other quasi's). Thus this limited corporate cash investing universe all but eliminates a scheme where it would be worthwhile for a company to borrow cheap debt and then try to earn a spread by reinvesting that $ into risky securities or bitcoins or sports bets or R&D into flying cars etc.

[1] https://fred.stlouisfed.org/search?st=corporate+debt

[2] https://www.sifma.org/resources/archive/research/

[3] https://www.bloomberg.com/news/articles/2017-08-15/amazon-is...

[4] https://www.sec.gov/Archives/edgar/data/1018724/000119312517...

[5] https://en.wikipedia.org/wiki/Interest_rate_parity

[6] https://www.ecb.europa.eu/stats/financial_markets_and_intere...

[7] https://www.ecb.europa.eu/stats/html/index.en.html

[8] https://fred.stlouisfed.org/series/BAMLHE00EHYIOAS

Plus some companies (which will remain unnamed) are unable to use cash parked overseas for US purposes. But they can borrow against it at low interest rates for capital.

> are unable to use cash parked overseas for US purposes

"Unable."

Unable without incurring costs that make Bonds/Loans vastly more economical.

The US situation with repatriation is silly. I’m all for some level of taxation but right now it provides a negative incentive due to interest rates.

Apple and Microsoft, for those wondering.

really, it is any US company that earns substantial assets abroad that avoids (at present) repatriating the money. GE, Pfizer, Exxon Mobil, Citigroup, Google, Goldman Sachs, Walmart, McDonald's... they all have billions in offshore profits that they stash. But I imagine the Apple was the implied company.

What would be the economic implications of changing US policy to allow Stateside based companies to bring back overseas capital at treasury rate tax costs? I can't really fathom the implications here cause i'm not to privy on the economic theory behind this. Obviously their would be some kind of inflation due to all the money that would flood into the US market, but that would be good for all the corporations with stock, resulting in more money inflooding. But the investment in US based industry would receive a considerable boost, which would be potentially good. I don't really see a downside here, so why does current economic policy dictate such a punishing tax cost to bring overseas cash into the stateside market again?

Whole foods, as a stable profitable business, is the perfect kind of candidate for this type of funding. Not sure why this would ever be scrutinized.

Excuse my ignorance, 1.4 + US 10Yr Treasury is ~ 4%.

Would it not be better if they issue it with much shorter terms and hence the lower interest rate? ( Penny Saving not worth it ? )

Could the Interest or repayment be counted against Profit? ( Amazon dont like Profits anyway ). And hence no tax.

I have always wonder why some companies dont go Private. I suspect these bonds could only be issued with a Listed Stock Company? How much more, in terms of Interest rate would a private company have to paid to these bonds sold?

Interest rates are low and rising. Compared to any other type of financing, corporate bond issues just make sense.

Next stop: Amazon offering their own credit card, outside of MasterCard/visa, with a bond offering being a way to dip into it.

Or insurance....

Bezos taking the Warren Buffet approach funding investments through insurance float?

That doesn't make any sense. Visa & MasterCard are the intermediaries, not the actual source of the cash. They sell their merchant network, and recreating it would be hard (not impossible, but a lot of work not in Amazon's wheelhouse).

Chase or Citi or Wells Fargo or whoever is the bank that needs reserves and fronts the cash to the merchant, and earns on the interest rate and fees and takes the risk on defaults.

"Earning the interest rate" is the key phrase here. With that and paying 2% in fees Amazon is leaving money on the table. Why not become the next Citi?

I'd bet that Amazon is paying far less than 2% in card processing fees.

I'd take the other side of that bet. Walmart & Costco get those rates because they can and have played hardball in negotiations with the network, they've shown a willingness to go without credit cards.

Amazon needs credit cards more than the credit card networks need Amazon.

Also, Amazon transactions are virtually all card not present transactions.

If they're made on-Amazon, and to someone with a 5-15 year Amazon profile, Prime, etc., shipped to home, etc., and with Amazon as the merchant (so they're going to handle CS rather than chargeback), yes, I'd take the bet that Amazon pays <2% for the Amazon Prime Chase Visa transactions for Prime customers on amazon.com. It's pretty low risk for MOTO compared to a card used at a random merchant online.

(Unless Amazon pays a higher fee but considers the 5% return to customers as some kind of "increase net spend on Amazon" benefit, but I think that's just Chase trying to kill Discover)

I've also wondered if the 5% back isn't something of an offset for the sales tax they've agreed to collect in some states. It doesn't totally cover the difference in high tax states versus ordering from someone else with free shipping with a 1-2% rebate card but it makes the delta small enough that many people may just default to Amazon.

Doesnt apply to me. Been with Amazon since 2005, prime 2 years now, bought a table that turned out to be scratched, open az claim, amazon said work with seller, seller was playing games for 2 months, no refund no return label for havy item that he promised is free to return, I went to bank and disputed the whole thing. Amazon didnt even blink, but it shows u your profile of a perfect amazon client might also ended up in dispute hitting their CB ratio.

Also as of this year Costco lost me as a client since they dont accept Amex anymore. To me that was enough.

Amazon's 2016 Annual Report lists $16.167B of "shipping costs" and $17.619B of "fulfillment costs", on a total of $135.987B of sales. Since shipping costs are included in fulfillment, this suggests to me that they paid at most $1.452B in payment processing fees, i.e., 1.06%.

But it's entirely possible that I'm reading the report wrong.

Amazon is 4 in every $10 spent online in the US. They have a scale similar to Costco/WM who only got deals for/to break exclusivity. Amazon pays less than 2%, much less depending on card type. Manages fraud to about 6BPS on transactional basis. But the concept of the running a CC is not out of the question, but the economics/complexity don't align yet...

The answer probably is: "Capital could be better spent on the core businesses" rather than that they can't.

Even Amazon with all its projects has limited attention it can spend. Becoming a bank isn't something you half-ass, which is why they'll just continue to do cobranded cards with companies who are already banks.

It's interesting to think of the bond market as the original kickstarter.

Also, the 40-year notes having ~4% interest seems like a pretty good deal for Amazon.

Isn't equity markets the kickstarter? You don't issue bonds to non-existent companies with a fresh idea. That's what equity (usually via VC) does.

The stock market actually does the opposite of raise money. More dollars go out with share buybacks than are raised with new issues. It's mostly used for equity owners to cash out. I mean, hardly anyone would own stocks if they weren't liquid, so it does aid investment, but not directly.

I should have better qualified my statement as Kickstarter is used by both new companies and ones with established revenue.

> Kickstarter is used by both new companies and ones with established revenue

Sure, Kickstarter is like the capital markets in that both give companies money in exchange for promises, not goods or services. Not sure how useful this analogy is.

Reply on news.ycombinator.com