My considerations about "holding cash" in Should you "be in cash" were entirely focused on what an investor should to about it -- my ideas are very different about what a firm (that the investor part-owns: remember that I consider it crucial to think about investing as owning a part of a firm!) should do about it are very different.
The Modigliani-Miller theorem shows that (under totally idealized conditions) it makes no difference at all what a firm does about its finances -- retain earnings or distribute them in whatever way (dividends, share buy-backs, &c), incur debts or sell more equity, whatever -- the theorem is often known as the capital structure irrelevance principle, exactly because it shows that, under certain assumptions, all of these details about a firm's capital structure and financing are just that, irrelevant.
However, much as the maths in question may fascinate me (and much as I may be personally biased towards this theorem by Modigliani being Italian;-), I consider the beautiful theoretical result in question pretty much irrelevant to the real world -- after all, among those "certain assumptions", it assumes efficient markets (which I consider a myth), no taxes (don't you just wish;-), no asymmetric information (ha!-), no agent-principal issues, and more.
In real life, a company's decision to incur debt (and in what form -- long term vs short, bonds vs bridge loans, intermediate forms now a bit out of favor like preferred stocks and convertible bonds) or hoard cash or both, and of whether, how much, and in what form, return "extra" cash to shareholders, does make a huge difference.
In particular: capital markets can and do "just freeze up", as the recent "great recession" amply proved -- making it risky to hold (especially) short-term debt in excess of the liquid assets available to repay it if needed. Acquisition opportunities do emerge, and being able to make a buy-out offer that's mostly or all cash (as opposed to acquiring-company shares) does make a difference (especially in credit-straitened times) to the chance that the offer will be successful. These considerations push many companies to hold a veritable treasure-hoard of cash and cash-equivalents (to the point where I believe that recently many otherwise-excellent companies, today, have gone overboard in this respect, holding far too much cash in their coffers -- mostly an understandable over-reaction to what the credit markets did over the last three years, I believe).
On the other hand, taxes are, of course, a reality, and have long encouraged companies (weirdly, in my opinion) to hold debt -- interest charges can be deduced from taxable earnings, while dividends, owners' capital gains, &c, cannot -- so, if a company needs a certain amount X of working capital to keep producing its earnings, and people who can supply it with capital require a certain remuneration (yield -- including interests for lenders, and dividends and capital appreciation for shareholders), it's cheaper for a certain fraction of X to be debt (up to the point where potential lenders become nervous about the company being over-leveraged and start demanding higher interest rates) than it is for it to be all equity.
The optimal debt-to-equity ratio, BTW, depends on the specific firm (and, in large measure, on what industry the firm is in): regulated utilities, in particular, have steady, predictable, "boring" income flows (and "boring" is a good thing for a lender), so they may leverage up with very substantial amounts of debt and still pay decent rates of interest -- while their lack of substantial growth prospects makes potential shareholder demand higher dividend yields than they do from firms in other industries (shareholders in utilities aren't looking for prospect of big capital gains). So, the assessment of whether a certain firm (that you're considering investing in) may be over-leveraged (a big risk to shareholders) or under-leveraged (holding too much cash -- not a risk, but depressing to prospective returns on equity), must depend on analyzing these factors.
And then sometimes there's a bubble in bond markets (we've been going through one of those for a while now!), and solid companies (ones where nobody's worrying about their going bankrupt) can sell bonds at ridiculously low interest rates (like MSFT's recent bonds that pay less than 1% interest!) -- when (for a company with no real risk of bankruptcy, and still some hope of future growth though quite possibly well beyond the "early tumultuous growth" phase of its life cycle) bonds pay much less yield in interest than stocks do in dividends, it's no doubt prudent and responsible for management to "take the near-free cash" by selling such bonds (as long as the resulting increase in the cash hoard is returned to shareholders in some way, or at least used responsibly to fund reasonably-profitable growth, rather than held entirely in sterile cash or squandered in wasteful acquisitions, perks, &c).
Of course, that parenthetical "as long as" condition can be quite a problem (and that's where agent-principal problems can emerge...), but, that's a discussion well worth its own post. The gist of this one is, rather, that whatever the theory (Modigliani and Miller's theorem) has to say about it, there are proper, prudent and profitable ranges of debt/equity ratio for a firm (depending on its industry sector, credit-market conditions and prospects, and other such considerations) and a prospective investor (or lender) would be well advised to take them into due consideration (carefully avoiding "gut" reactions about firms having "too much" debt, or "too much" cash -- either condition is quite possible, and occasionally even both are observed together!, but each such case must be weighed on its individual merits).
In future posts I plan to discuss acquisitions, and the issue of how (as well as whether and how much;-) to return "excess cash" to shareholders rather than hoard it (and/or use it to pay down some debt prematurely) -- dividends (regular and special) and stock repurchases. All of these issues are enmeshed with potential problems that can be couched in principal-agent terms (i.e., potential conflicts of interests and incentives between the firm's owners and its management), and indeed one could say that the agency dilemma is at or near the core of each and every one of them...