Friday, April 10, 2009

Thoughts on rights issue

I actually wrote this almost 2 months ago. While I've circulated to some close friends to bounce off some ideas, I decided that I should only post this after the successful completion of my employer's rights issue. Realistically, the material below is not specific to any particular rights issue (though I did refer to examples of specific ones to make my point), rather it's a collection of some thoughts on how a shareholder should interpret/evaluate a rights issue, and it is my contention that the media at large has done the public a big disservice by possibly presenting a wrong framework to look at the whole issue. Having said that, I must confess that I'm really no expert and it's totally possible that my points here below are not relevant or worse, wrong in its conclusions. Indeed i'd be more than happy to hear your views on how I might have misinterpreted the mechanism or screwed up the analysis.

Finally, I've posted 2 models on google spreadsheet, one done by me, and a follow-up done by my friend Eli. If you're the number-crunching kind, I suggest you dive straight to them. The models are fairly straight forward.

So, here goes, a long post on the topic of Rights Issue:

When is discount a discount?

Let's start with something fairly recent. CapitaLand announced a 1-for-2 renounceable rights offer for an equity raising of ~S$1.84b, at S$1.30 per new share or 45% discount to the trading price prior to announcement (S$2.36). My first impression upon learning the news was that the offer price was at a steep discount, and as an existing shareholder, I thought that it's a bargain that I shouldn't miss. Perhaps I could "average down" my per share cost? How about any dilution effect? Are we transferring wealth from existing shareholders who're not participating in this equity raise to those new shareholders who's getting the shares at a discount?

I was not alone in this line of thinking. The next day, BT runs an article on the rights offer with the following line: "...another element that caught most analysts by surprise was the steep discounts at which the rights issues are being done..." while quoting another analyst saying that "CapitaLand and CMT could be pricing the rights issues lower to entice their shareholders to take up their allotments in the current weak market."

Then the following day, BT runs another article, making a similar point, "...This has led some to suggest that such hefty discounts - disbursed by two of Singapore's biggest companies, no less - may become the benchmark for subsequent issues. If DBS and CapitaLand are pricing at 45% discount, can much smaller companies ask for anything higher? While there are many factors influencing pricing, such as the number of rights shares against the number of existing shares held and size of rights issue, the market is bracing itself for more deeply discounted deals to follow. This will be bad news for weaker firms that are trying to raise cash from shareholders, now that the banks have turned their backs on lending. Without deep discounts, rights issues may fail if shareholders shun them..."

However, on second thoughts, this "discount" somehow doesn't feel right. Firstly, share price is a function of the shares outstanding, but the number of shares outstanding can be entirely arbitrary. The only figure with "real" economic meaning is the market cap (although strictly speaking, even the market cap is just a stretch of imagination of what very few people, at a particular point in time, think the common equity portion is worth and we just extrapolate that as what all the other non-participating must also think; certainly an intellectually convenient shortcut to quantify worth, but for our purpose can be said to be more "real" than share price or shares outstanding). Indeed, comparing Berkshire's A share price (~US$90,000) and Microsoft share price (~US$19) has no meaning while comparing their mkt caps (US$140b vs US$170b) gives real insight.

In other words, the rights issue ratio, with which the ending outstanding shares are calculated, can be adjusted to achieve different "discount" offer price, while ensuring that the amount of new equity raised is the same. If so, isn't this "discount" an illusion? In fact, could the issue ratio be adjusted to produce a "premium" offer price instead of a “discount” price? Will my ending economic interest as a shareholder be enhanced or eroded due to such an offer? I figured it'd be easier to just do a quick model to find out.

The analysis

Assume a hypothetical company with 36 shares outstanding originally (see below). The current traded price is $2.00, giving a mkt cap of $72.00. Also assume that there’re 3 existing shareholders, each holding 12 shares or one-third of the equity (although they each have a separate average cost of investment, with investor A buying in at the highest at $5, and investor C buying at $2, the current price). Note that I’ve also computed the cost of stock for each investor per 100 parts of the stock (eg. Investor A: $60/33%/10 = $1.80). I’d explain the significance of this additional measure later.

Now suppose the company announce a 1-for-2 rights offer (similar to CL’s arrangement; see case 1 below) at an offer price of $1.00 per new shares, or 50% below the last traded price of $2.00. Assuming full take-up, the company will be able to raise a total of $18 with a Theoretical Ex-Trading Price (TETP) of $1.67 and pushing the mkt cap from $72 to $90. Here, we show that both investor A & B fully take up their respective rights, hence maintaining their proportional share of the stock (at 33%). Doing so will naturally lower their cost of investment on a per share basis (investor A from $5.00 to $3.67 & investor B from $2.00 to $1.67). However, such comparison is nonsensical as the very yardstick used in this measurement (ie. number of shares outstanding) is different before and after the equity raising. This is like measuring a shoe lace with a ruler based on centimeters standards and measuring a slightly longer shoe lace with another ruler based on inches and then concluding the shorter shoe lace is actually longer because of the nominal value in cm.

Instead, the more consistent way to calculate the cost of investment in a stock is to use a fixed base, in this case I’ve chosen to split up the stock into 100 parts. As such, we can see below that the cost of stock for Investor A has increased from $1.80 to $1.98 and for Investor B $0.72 to $0.90 after the rights offer. This makes sense because while both investors A & B maintain the same proportion of the stock (both 33%), the pie (or worth of the company) has become larger so naturally the cost of investment on a per 100 part basis would have increased.
Now what if an existing shareholder declines to participate with this equity raising? How would his economic worth change after such an exercise? If we only think about the value of the stock, it would seem that his economic interest would have been diluted. However, the shareholder has a second recourse, which lies in the value of the rights he’s been conferred.

Here, we show that investor C does not participate in the equity raising, so he’d have maintained his original number of shares (12 shares). What should he do with his 6 “rights” then? These “rights” are essentially call options, and call options have 2 sources of value, its intrinsic value if it’s in-the-money and its time value (for the potential of being more in-the-money before maturity). We can think of the offer price of $1.00 as the strike price of the option, and if we assume for the moment that the TETP of $1.67 will indeed be the ex-trading price, then this “right” will have an intrinsic value of $0.67 (we’d ignore the time value for now; in any case, it’d be a very small value due to the typical short trading window associated with rights). Then, investor D, someone who has not prior shareholding in the company, comes along and if he also agrees that the ex-trading price will be $1.67, he’d be willing to purchase these rights at $0.67 apiece from investor C. The end result is that investor C’s stake in the company would have been diluted from 33% to 22%, but he’d have realized a cash gain from the proceeds of the rights sale, with the net effect that his economic worth would have maintained the same at $24 although they’re now being represented in different assets. Previous, he had 12 shares at $2.00 (giving $24), now he has 12 shares at $1.67 and cash of $4.00 (also giving $24). One interesting outcome is that his ending cost of stock (net of cash proceeds from rights sales) on a per 100 parts basis would have remained at $0.36 (ie. (12 x $1 - $4)/22%/100=$0.36 ). Why is this the case?

Assume then instead of a 1-for-2 rights offer, the company announces a 1-for-1 offer. In order to raise the same amount of equity ($18), the offer price for such shares would be $0.50 (instead of $1.00 previously). The ending mkt cap will be the same, at $90. Looking at both investor A & B (who both participate fully in offer and hence maintaining 33% stake each), their per share cost of investment would appear to have reduced (eg. Investor A from $5.00 to $2.75) but on a per 100 parts basis, its ending cost would be $1.98, the same as it were under a 1-for-2 offer. In other words, the existing shareholders who take part fully in such issue should have been indifferent between different “discount” in offer price or rights issue ratio.

How about investor C, the shareholder opting to sell his rights? He would have received $9 in this case if he gives up all his “rights to pay up” and his stake would have been reduced from 33% to 17% correspondingly while maintaining his original economic worth of $24. On the other hand, his cost of stock on a per 100 parts basis would have been $0.18 (ie. (12 x $1 -$9)/17%/100 = $0.18) versus the original $0.36. This is true “averaging down” of cost since this is based on a fixed scale (ie. 100 parts) instead of the usual averaging down when most think in terms of cost per share.

Let’s take a closer look to see how this reduction in cost of stock (per 100 parts) is achieved. Table 4 shows Investor C’s possible profile under both 1-for-2 (blue) and 1-for-1 (red) offer scenarios as he adjusts his ending stake in the company. He can vary his stake by opting to participate partially in the rights offer (selling some rights while exercising the rest himself). We can see here that the 1-for-2 scheme only allows Investor C to “sell down” his stake to ~22%, while 1-for-1 scheme allows him to go down to ~17%. Hence, a 1-for-1 issue actually allows greater scope for true “averaging down” (in per 100 parts basis) given the nominal discount in offer price is less than that of the 1-for-2 issue. Naturally, if investor C chooses to fully take up the rights issue, his “net worth” (accounting for the stock value and cash value) would have increased to $30 like investor A & B, but if he’s willing to exchange his asset mix from one form (in this company’s stock) to another (cash proceeds from rights sales), he might be able to do some true “averaging down”. However, I have to caution that such “averaging down” only works for the investor whose original cost of investment (on a per share basis) is higher than the offer price.

As a final note on the “discount” price, I’d like to add that in fact there should be no realistic way for the right issue to be priced at a “premium” and hence it follows that all rights issue would have to be offered at discount. One way to think of it is that if it’s priced at a “premium” to current trading price, the rights (or essentially options) will basically have a negative intrinsic value! Unless there’s legitimate reason to believe the trading price will shoot up after the announcement of the rights issue, the negative intrinsic value basically means that the shareholder will be handed a liability instead of an asset if given such rights.

Some other considerations

There’re a few more important aspects surrounding a right issue that I’d like to point out briefly (yes, this post is getting too long). First is the treatment of odd lots. This arises when the issue ratio is in some “funny” ratio whereby it’d be rather rare for the existing shareholders (especially retail investors) to convert his shares nicely into whole-number lots. I’m not sure if there’s any convention in dealing with this “odd lots” trading issue. Quite likely, there’d be a limited trading period for such lots in the market so that the lots can be made “whole” as much as possible through market transactions. I haven’t thought too deeply into this yet but I suppose the value of such “odd lots” would be slightly depressed compared to full lots simply because we have a case of motivated sellers. Just think of whether you would prefer a one-dollar coin versus 100 one-cent coins that come with expiry date!

Secondly, what if the rights issue is underwritten by 3rd parties? Essentially, the underwriters will guarantee to take up the issue if the rights are not fully taken up. But there’s a catch, these 3rd parties will pay for the issues at the discounted price, without needing to pay for the rights to acquire the new shares, because the rights have expired! This gives rise to an interesting dilemma. Basically, existing shareholders are offered the rights to buy up new shares. Now these rights have intrinsic values but they are offered to existing shareholders for free simply due to their position as co-owner of the business. Question is, why should there be an expiry date to these options and if there must be one, how long should the option tenure be in order to be considered fair? Obviously, the management calls for rights issue because it needs/wants capital injection, and presumably they want it sooner rather than later. The process admittedly shouldn’t be delayed unnecessarily but is a fair option tenure a week? A month? This is open to debate.

Final thoughts

At the end of the day, the crucial question a shareholder should ask before committing to rights issue is to decide if the company can use this new funding to produce ROIC sufficiently higher than the firm’s new WACC. If this new funding is meant for new growth opportunity, then possibly a rights issue makes sense. But if the new funding is meant to effect permanent change to the firm’s capital structure (eg. gearing), then the shareholder needs to decide whether he’d be happy if the lower return is sufficiently higher than the lower cost of capital. What certainly shouldn’t be the shareholders’ consideration (if my analysis above is correct, which I confess I’m not 100% certain) is that he shouldn’t miss out a good chance to buy shares at “discount” to cost “average down”.

Which is why I get frustrated to read things like this in the press: “Funds will sell ahead of rights issue because the rights are at a discount and cheaper than the mother share”, said Peter Ng, CEO of boutique fund on CMT rights issue….Issuing new shares at steep discount to market price is seen as one way to make these cash calls more palatable to shareholders.” – The Edge 16 Feb 2009 issue.

Indeed, I’ve always had a big problem with such misplaced trust in seemingly knowledgeable source. Journalists, economists, even doctors (just google “iatrogenics”) are all common culprits. I shudder at the thought of the many disastrous decisions made based on such over-simplification and reliance on the “obvious” reasoning. I maybe too disparaging to conventional “wisdom” (and I know I can get on people’s nerves because I’m always so skeptical), but I guess I just hate “bull shit”, especially when the fellow don’t even realize it and brand it as “advice” instead.

No comments: