The warrant had provided Berkshire Hathaway the right to purchase 43,478,260 shares of Goldman Sachs' common stock, par value $0.01 per share, at an exercise price of $115 at any time until October 1, 2013. Under the amended agreement, Goldman Sachs will deliver to Berkshire Hathaway the number of shares of common stock equal in value to the difference between the average closing price over the 10 trading days preceding October 1, 2013 and the exercise price of $115 multiplied by the number of shares of common stock covered by the warrant (43,478,260).
So the warrant agreement has been converted from cash settlement to net share settlement. Some things worth noting:
- The amended deal eliminates some risk of dilution for Goldman. Let's say Goldman's common stock trades near current levels next fall. Well, under the new terms, Berkshire ends up with a much smaller number of shares (roughly a fifth as many). The end result is a bit less than 2 percent of additional shares outstanding. This will likely place Berkshire among the top ten owners of Goldman shares this fall.
- Under the original scenario, Berkshire would have been able to purchase ~ 43.5 million shares of Goldman at $ 115. That would have, instead, resulted in roughly 9 percent of additional shares outstanding.
- The benefit for Berkshire is that it does not have to lay out a big chunk of change upfront in order to exercise the warrants. Naturally, as a result, the company also ends up with far less exposure to Goldman's common stock.
(Though Warren Buffett could just sell a portion to bring it in line with the ownership level he considers appropriate.)
- If Goldman's shares happen to trade near the recent price (~ $ 145 per share) over the 10 trading days prior to October 1, 2013, Berkshire would end up with ~ 9 million shares worth roughly $ 1.3 billion.* In the prior scenario, Berkshire would be paying $ 5.0 billion to get more like $ 6.3 billion (the same net amount). Basically, instead of writing a check to buy 43.5 million shares at $ 115 per share, Berkshire is getting compensated in stock for the difference between Goldman's market price this fall and the $ 115 exercise price.
A similar result actually could be accomplished without an amendment.**
This should work out fine but what if a big move in the stock in either direction occurs? The strike price inherently provides substantial leverage, so it wouldn't take much of a move in either direction to make a meaningful difference in value. Also, in the original deal, Berkshire had the right to act at any time. In this deal, what happens to be the average closing price over ten trading days determines the value. Seems less than ideal, but there are many trade-offs to consider here.
So, yes, if Goldman's stock is selling near the current price during the crucial 10 preceding trading days, Berkshire will get the ~ 9 million shares worth roughly $ 1.3 billion.
Yet, unless there's something in the fine print, if the stock drops to an average closing price of $ 115 -- just over 20% lower than now -- or less on those 10 trading days, Berkshire would end up with no shares and the value of this amended deal will be zero. Unfortunately for Berkshire, that'd be true (and quite annoying) if Goldman's share price recovered shortly after that.
Now, if the opposite were to happen -- a roughly 20 % increase in Goldman's stock price -- it would clearly be a nice benefit to Berkshire.
(Berkshire would end up with nearly 15 million shares of Goldman instead of 9 million.)
A relatively minor decrease in the stock price results in Berkshire getting meaningfully fewer Goldman shares (or maybe even none). A relatively small increase in the stock results in Berkshire getting quite a nice bump in additional Goldman shares.
In a calm market this dynamic probably isn't a big deal. Still, agreeing this far in advance seems to potentially create a rather unpredictable and wide range of outcomes.
Who knows what the world might look like this fall.
Both companies naturally understand this well but clearly have decided the benefits of the new terms outweigh the negatives.
* Here's the math assuming a $ 145 average 10 day closing price: (43,478,260*(Avg 10 Day Closing Price - Exercise Price))/Market Price = 43,478,260*($ 145-$ 115)/$ 145 = $ 8.996 million shares
** Alternatively, this fall (using the same average 10 day closing price) Berkshire Hathaway could just pay $5 billion to Goldman Sachs, receive the $6.3 billion in stock, then sell $5 billion (back to the market or Goldman) and end up with the same net $ 1.3 billion in stock. Goldman could also take the $5 billion paid by Berkshire in cash (for the $6.3 billion worth of stock), then just buy $5 billion back (from the market or Berkshire) using that cash. The mechanics of doing this may not be terribly difficult, but it's also not necessarily all that easy to buy back $ 5 billion of stock (at least not very quickly). There's the transaction costs. There's dealing with potentially unpredictable price action (never mind the fact that $ 5 billion of buying is enough to move the stock). This amendment's approach makes the whole process quite a bit cleaner for both parties. In the amended form, Berkshire doesn't have to write a big check upfront and doesn't end up exposed to so much Goldman stock. (It's worth noting, under the original deal, that Berkshire apparently wouldn't be able to sell the $ 6.3 billion in stock right away. So if Buffett wanted to reduce Berkshire's exposure there'd at least be a lag before it was possible to do so.) In the new deal, Goldman also doesn't have to deal with buying back the stock to reverse the dilution. With the original deal, consider the possibility that regulators (and possibly politicians) might just pressure (or require) Goldman to hold onto at least some of that $ 5 billion in cash received from Berkshire (from purchasing the warrants). I'm guessing there's more than a few who wouldn't mind if at least some of the bigger banks had additional liquidity for a rainy day. Too-big-to-fail is still rightfully front and center in the minds of many. Still, short of forcing the bank to hold onto the capital instead of using it buying back the stock, Goldman would end up with the same capital levels when it's all said and done under both scenarios (even if the mechanics are quite different). It's worth pointing out that the regulatory filing by Goldman states that the new deal becomes effective only if the "Board of Governors of the Federal Reserve System has approved or has stated that it has no objection to the net share settlement of the Warrant."