Over the past several years, Micron Technology (NASDAQ:MU) has regularly issued convertible debt and assured shareholders they are protected from dilution by "capped calls" which were purchased as insurance. (Only the "sub basement" of our towering stock price rise was so insured.) Shareholders were told the latest issuances were required to facilitate the purchase of Elpida, completed almost a year ago. Shareholders were not told that the latest issuances needn't have been done, since Elpida had almost enough cash on its own balance sheet as of closing to pay for the entire purchase. Shareholders continue to be materially misled by this chart, which appeared in the latest quarterly earnings call materials:
Through the miracles of non-GAAP accounting, Micron implies that the dilution is "only" 130 million shares, when in fact, the number is larger.
So what is the dilution from these things? In its last Form 10-Q, Micron showed $2.503 billion of convertible debt outstanding. This convertible debt has conversion prices so deeply in-the-money that this little table, on page 17 of the 10Q, shows conversion value of $2.662 billion in excess of the par value of the bonds.
In Note 2, it states this Conversion Value in Excess of Principal was calculated at $28.58 per share. Marking that to market, with the closing price on July 23, 2014 of $33.68, one gets an additional $5.1 on the underlying 179 million shares, or an additional $0.913 billion. Adding the marked to market conversion value to the principal, one gets about $6.029 billion for the size of this pothole in our balance sheet.
The actual share dilution from these converts is 179 million shares, or about 17%, and is NOT the 130 million shares implied in the previous slide.
Previous articles. In previous articles (here and here), I argued that these converts shouldn't have been issued and were improperly hedged. While I still feel that way, I concede defeat on winning new adherents to this point of view. Where I did have some success was in gaining some support that they should be redeemed ASAP, before the stock climbed further and created a larger liability (a larger liability from an improved stock price is exactly what has happened). The company has been chipping away at repurchasing the converts in the market.
So it was a welcome 8-K which detailed that the company intends to issue straight debt and use proceeds to retire the 2031 tranche by the end of August. (Apparently, demand for the debt was such that the total issued was $1.15 billion, greater than the $750 mm originally anticipated. Again good news that our straight debt is in demand!) That 8-K also detailed the repurchase of $58 mm of the 2032 tranches for $197 mm cash.
Put it in Perspective. Issuing $1.15 billion of straight debt to redeem more of these awful converts is nothing to be sneezed at. I've been one of the harshest critics of these converts, and applaud the effort. Still, $1.15 billion is only 18% of the total "as converted" value, as shown above. I believe the company should be congratulated, but could and should be doing more.
Here's a partial listing of what I'd like to see happen to repair our balance sheet, which CFO Ron Foster has so grievously injured and which he is only slow to repair:
- Since our straight debt seems to be in demand, we should issue more of it to deal with more of the converts. If they are not callable or purchasable in the market, then we should buy in stock to settle eventual conversions.
- Once we have done #1, we should declare a small cash dividend. This will eliminate some of the short interest in our stock.
- We should adopt and stick to some very concrete steps on returning cash to shareholders. SanDisk (SNDK) has stated that it will return 100% of free cash flow to shareholders via buybacks and dividends. Intel (NASDAQ:INTC) has recently added $20 billion to its already impressive buyback program.
If we do these things and if we continue to perform, as I believe we will, the company will be rewarded with a higher price-earnings multiple (or in a tip of the hat to cash flow guys, a lower discount rate). There is a reason why Google Finance tells me INTC's PE is 17.12x, SNDK's is 19.12x, and MU's is only 11.23x.
Conclusions. I applaud Micron finally getting after the converts in a big way. I wish it was a bigger way.
Around the time of the last earnings call, I closed out a large options position numbering tens of thousands of contracts, covering hundreds of thousands of shares. I continued to hold a small stock position. I was disillusioned that management was treating the convert issue as a high priority, as they said they were. I was disillusioned by Mark Durcan's remarks at a Bernstein analyst presentation that we were 2 to 6 quarters behind in our NAND offerings. The issuance of $1.15 billion in debt shows the company is FINALLY beginning to take serious steps on the converts. The passage of time and SanDisk's latest earnings call show that our position in NAND may be alright. In any event, we have now survived 1 of the "2 to 6 quarters" that Durcan says are needed to right our NAND wrongs. So... I have purchased several thousand contracts of the MU 2015 January 35 calls and wait anxiously for more positive news, which I think is on the way.
Disclosure: The author is long MU, SNDK, INTC. The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.