Investors must learn to accept that in an unbalanced world reliant on the printing press, everything is risky, including holding cash.

—George Karahalios, The Gloom, Boom & Doom Report, August 29, 2014

In May 2007, we published Why Not Hold Cash? in which we argued that for most investors, holding cash or derivatives as a tactical hedge against broadly expensive asset classes did not make sense due to the complexities and behavioral issues surrounding such allocations, and encouraged most investors to stick with the policy targets in their diversified portfolios, which, as we wrote, “usually provide greater protection in a market downturn.” What we missed in advising not to hold cash as a tactical hedge is encapsulated in a footnote from that paper, in which we stated that the one exception to the rule—i.e., when diversified portfolios do not help—“has been during extreme, crisis-type events … when correlations among all asset classes (except ‘safe havens’ such as US Treasuries and gold) have generally moved toward one.” Of course, as everyone now knows, just such a crisis was coming.

For a discussion on the similarities and differences between the 2007 environment and today’s, please see Eric Winig et al., “Echoes of 2007?,” Cambridge Associates Research Note, July 2014.

Seven years later—a remarkably short period of time, it must be said, for markets to have subsequently plunged and re-ascended to worrisome levels—investors find themselves in a similar, if not worse, predicament. Asset classes are broadly expensive, and in contrast to 2007, the global economy is sputtering, with a level of geopolitical disorder not seen since the late 1970s. Central bank and government intervention has also muddied the waters of dispassionate financial analysis, as it is impossible to disentangle the effects of such policies from organic economic activity and corporate profits. The enormous sums disbursed by governments since the 2008 crisis, some large fraction of which has likely shown up on corporations’ bottom line, make it possible that most asset classes are even more expensive than they appear. The risk of an exogenous economic shock also seems higher, perhaps much higher, than in the recent past.

Further complicating the question of whether to hold cash is that while a decent return on cash could actually be earned in 2007 (the Federal Funds rate, for example, was 5.25% when we wrote the paper, compared to CPI at 2.7%), this option is notably absent today, with the Fed Funds rate at 25 bps and the CPI at 1.7%. And, whereas in 2007 it was unclear how far central bankers would be willing to go to support markets, today there is no ambiguity, as best summed up by Mario Draghi’s 2012 “whatever it takes” proclamation.

The juxtaposition of these realities suggests that the next crash may be (a) further in the future than many believe, and (b) worse than almost anyone expects when it finally arrives.

Taking into consideration all the challenges of the current market environment, including that markets may well continue to rally^[1]While this seems less likely given recent declines, it is worth noting that the market has recovered strongly from similar drawdowns over the past few years, and the speed with which central bankers … Continue reading investors today should consider holding cash—or preferably, increase sovereign bond allocations—given the extraordinary number and scope of current unknowns. Such an action remains subject to all the complexities and behavioral challenges we outlined in 2007, not the least of which are accepting foregone gains and the potential to underperform peers for some period of time.

The Valuation Conundrum

While asset classes are broadly expensive today, they do not, at first glance, exhibit the same level of overvaluation as in 2007. For example, according to our valuation matrix—whose methodology for most asset classes has remained relatively consistent over time, although the number of asset classes has grown^[2]An analysis of a constant universe over this period showed a similar pattern.—half of the asset classes we cover today are fairly valued (34 of 67), with 6% undervalued, one-third overvalued, and 10% very overvalued (Figure 1). This compares quite favorably to May 2007, when 30% were fairly valued, nothing was undervalued, a whopping 60% (31 of 52) were overvalued, and 10% were seen as very overvalued.

However, we would caution against drawing too much comfort from such statistics. As noted above, it has become increasingly difficult to fundamentally value assets as central bank and government manipulation of markets has become not just accepted, but also expected. Thus, we are left scratching our heads when trying, for example, to assess what is “sustainable” for US corporate profits given that recent per share growth has been driven to a large degree by zero interest rates, which have not only pushed down finance costs, but also boosted debt-financed share buybacks and encouraged consumption at the expense of saving. All this has boosted profit margins, even as topline revenue growth has remained weak; further, the lack of investor differentiation among equities (as more and more simply buy “the market”) has driven up prices across the board. As a result, while our composite normalized price-earnings measure looks expensive but is nowhere near its 2000 peak (Figure 2), the median price-to-sales ratio—which adjusts both for historically high margins and the extent to which valuations are skewed by larger firms—is not only above 2007 and 2000 levels, but at its highest level ever in our post-1978 dataset (Figure 3).

In short, while at first blush valuations look better than in 2007, we view this with a jaundiced eye given how much of recent economic and financial activity has been driven by central bank activities. We have no doubt they plan to continue such actions as long as they deem it “necessary,” but are reminded of the late Herb Stein’s famous admonition that something that can’t go on forever … won’t. Current valuations—for virtually all asset classes—would look a good deal worse if such support were withdrawn.

What About Opportunity Cost?

The biggest obstacle for most investors to holding cash is foregone gains. Indeed, this is often true even for investors that go into such trades with eyes wide open, i.e., with the stated goal of loss prevention at the expense of additional gains. Unfortunately there is no easy answer to this behavioral issue; it is remarkably predictable that (initially cautious) investors’ risk aversion will decline as markets rise and their returns lag those of peers. Thus, to resist the urge to chase returns in the event markets continue to run, investors that do raise cash or other “safe” assets should consider putting in place strong rules documenting not only the purpose of said allocation and how often it will be revisited, but perhaps even a range of “acceptable” foregone gains. Of course, investors could also stop comparing returns to peers, who often have different objectives, capabilities, and risk tolerances. We recognize that for many this is a bridge too far, but do encourage a conversation regarding this habit.

See Seth Hurwitz et al., “The Case for Investment-Grade Corporate Bonds,” CA US Market Commentary, November 2008.

The question of whether to hold cash always comes down to timing. One of the errors we made in 2007 was that, in our quest to be evenhanded about investor capability to implement such a strategy, we swung too far in the direction of assuming effectively no skill whatsoever. Thus, we presented the fact that portfolios with cash and/or out-of-the-money put options had underperformed over time as “proof” that such strategies are always losers (Figure 4). But of course no one actually invests this way—a cash allocation is nearly always intended as dry powder available to buy assets when they become cheaper. Indeed, it is this “optionality” of holding cash that makes it so valuable to asset holders over time. To take one obvious example, an investor that sold equities in May 2007, then used the cash to buy corporate bonds in early 2009 (as we and others recommended), would have come out far ahead.

All that said, we do not mean to downplay the difficulty in getting the timing right. While the question asked in our 2007 paper seems well-timed in retrospect, we could have made a very similar argument in 2006; one of the hallmarks of bubbles is that they tend to be more powerful and durable than most expect. As the saying goes, sometimes being early is indistinguishable from being wrong.

Finally, we would be remiss to ignore the possibility that the next crash will be so bad as to render virtually all defensive strategies moot (e.g., a synchronized collapse in fiat currencies). While certainly not our base case, the fact remains that no one has any idea how the current central bank experiment (Figure 5) will play out, and given the record of centrally managed economies some caution would seem to be in order. Put simply, there is some non-zero chance the global economic system will need to go through a “reset” that includes a revaluing of developed markets currencies. Somewhat paradoxically, given the near impossibility of protecting against such an event, the most rational response for one that expected this outcome might well be to throw caution to the wind and, in the infamous words of Chuck Prince, dance while the music’s playing. As the saying goes, never bet on the end of the world, since it only comes once … and how will you get paid?

But Seriously … Cash??

At this juncture, rather than raising true cash, we would instead advise increasing the allocation to sovereign bonds, for several reasons. First and foremost, sovereigns offer true diversification for equity-dominant portfolios, with an expectation they will pay off when most other assets are sinking. A big part of what we missed in 2007 is that the vast majority of assets in most investors’ portfolios are risk assets. While such assets can and do perform differently during most periods, correlations trend toward one during extreme events (e.g., the 1998 Russian debt crisis and of course the 2008 global financial crisis). In the event that central banks prove unable to stop a global slide toward deflation, sovereigns would likely prove one of the few appreciating assets.

Second, the rub against sovereigns in recent years has been low yields. Sub-3% yields on US Treasuries—to say nothing of 1.5% French bonds and 1% German bunds—are hardly the stuff of investor dreams. Still, the obvious counterexample is Japan, where yields not only continued to decline in the face of ever-more-strident warnings that they could go no lower, but are even now plumbing new lows nearly a quarter-century after the country’s asset bubble burst. Indeed, the outperformance of Japanese government bonds over cash since 1990—not to mention equities—is nothing short of remarkable (Figure 6). Even more pertinent for current investors is the fact that Japanese government bonds have handily outperformed over the past 15 years, posting a real average annual compound return of 2% despite a starting nominal yield (on August 31, 1999) of 1.92%.

For a detailed discussion of this topic, please see Stephen Saint-Leger et al., “How Far Will US Rates Rise in the Next Cycle?,” Cambridge Associates Research Note, September 2014.

Finally, the verbal commitment of central banks to keep rates low—coupled with the fact that what economic growth does exist (e.g., US and UK housing, and US auto sales) is highly dependent on low rates—strongly suggests rates will stay low for some time.^[3]We generally advise investors to hold bonds denominated in their home currency, since any benefits gained by international diversification are offset by the fact that the bulk of the spending being … Continue reading Simply put, in the event markets collapse and one’s cash holdings “pay off,” we would expect sovereigns to perform even better, and the risk of rates spiking higher anytime soon seems relatively low. Even in the unlikely event that economic growth takes off and drags rates higher, the losses suffered on sovereigns would almost certainly be outweighed by the gains on risk assets that make up the bulk of most portfolios. Ironically, the best time to hold cash (as opposed to sovereigns or risk assets) would be during a rising inflation period—with or without strong economic growth—that would hurt sovereigns and risk assets simultaneously.

While investors will of course differ in how much cash or Treasuries they choose to hold, our base recommendation (for those with allowable ranges for asset classes) would be to at a minimum bring Treasuries above the mid-point of their fixed income/cash range.

Where Could We Be Wrong?

Well, pretty much everywhere, but the obvious issue is that of timing.^[4]As we wrote in our May 2007 piece, “while it is possible mid-2007 will prove (in hindsight) to have been a uniquely opportunistic time to sell richly valued assets and raise cash, the chances of … Continue reading The decision to hold cash is always a risk for investors that not only need to meet spending targets, but also measure performance against market benchmarks … and peers. Thus, while we believe prudence is the better part of valor at the moment, investors that simply cannot stand to underperform for the next year or two (or three!) should recognize that raising cash will “feel stupid” for some period of time. Among the few things we “know” at the moment are that central banks will keep the pedal to the floor, and (relatedly) any sustained downturn in risk assets will be met with even greater stimulation/intervention. It is worth noting that while our May 2007 paper seems remarkably well timed in retrospect (well, it would had we recommended holding cash), markets did not peak for another six months, and their ultimate denouement was nearly two years away; such timeframes always seem shorter in hindsight than they do in real time.

Another possibility is that risk assets are not in a bubble, but simply reflect that the United States is, as many currently believe, well on the way to a sustainable recovery that will allow the Federal Reserve to raise rates within the next year or so. (It is difficult to make a similar case for the Eurozone, which has the hallmarks of sliding into a deflationary morass à la Japan.) In this case cash might or might not prove valuable—given how well equities have performed with a moribund economy and accommodative Fed, it is reasonable to posit they might suffer under the opposite circumstances—but sovereigns would almost certainly suffer as rates rose. However, we will go on record as saying we regard this as among the unlikeliest scenarios on offer.

Conclusion

Should investors raise cash today? Given extended valuations, the growing disconnect between financial markets and underlying economies, and a long list of geopolitical trouble spots, the answer would seem to be yes. On the other hand, the verbal commitment of central banks to continue spiking the punchbowl—and investors’ near-unanimous belief in their power to levitate risk assets forever—is a powerful counterargument.

We recommend investors look to build “safe assets” as a precautionary measure—recognizing the possibility, or even the likelihood, that we are very early—but prefer long-dated sovereigns to zero-yielding cash, in the belief that price risk is more than offset by their diversification and potential capital appreciation benefits. Investors should also strongly consider putting in place specific language about the purpose of the allocation and how often it will be revisited, and perhaps even a range of acceptable foregone gains if markets continue to rise.

Contributors

Eric Winig, Managing Director
TJ Scavone, Investment Associate

Exhibit Notes

Figure 1. Distribution of Asset Class Valuation Calls by Cambridge Associates
Source: Cambridge Associates LLC.
Notes: Data on fundamental valuations do not provide forecasts of expected returns; they refl ect the vulnerability of a given asset class to disappointing economic and profit developments. Therefore, valuations may not necessarily correspond to short-term or even intermediate-term returns. For example, asset classes and investment strategies can be fairly valued yet still retain a negative outlook due to deteriorating fundamentals. Asset class and investment strategy valuations do not reflect currency valuations. The valuation category “dangerous bubble” did not exist in May 2007 or August 2014. No asset classes were “undervalued” as of May 2007, nor were any “very undervalued” as of August 2014, May 2007, or March 2000.

Figure 2. US Composite Price-Earnings Ratio
Sources: MSCI Inc. and Thomson Reuters Datastream. MSCI data provided “as is” without any express or implied warranties.
Notes: The composite normalized P/E ratio is calculated by dividing the inflation-adjusted index price by the simple average of three normalized earnings metrics: ten-year average real earnings (i.e., Shiller earnings), trend-line earnings, and return on equity (ROE)–adjusted earnings. The ROE-adjusted P/E ratio is the current trailing P/E ratio multiplied by the ratio of the current level of ROE to its historical norm. Shiller P/E is calculated by dividing the current price level by the ten-year average of inflation-adjusted earnings per share. Trend-line P/E ratios compare current stock prices to the level of earnings predicted by long-term real earnings growth based on a simple linear regression.

Figure 3. S&P 500 Median Price-to-Sales Ratio
Sources: Ned Davis Research, Inc. and Standard & Poor’s.

Figure 4. Cumulative Wealth Comparison of Various Portfolios
Sources: Bloomberg L.P., Barclays, BofA Merrill Lynch, Cambridge Associates LLC, Citigroup Global Markets, Dow Jones Indexes, J.P. Morgan Securities, Inc., MSCI Inc., National Association of Real Estate Investment Trusts, National Council of Real Estate Investment Fiduciaries, Standard & Poor’s, Thomson Reuters Datastream, and Wilshire Associates, Inc. MSCI data provided “as is” without any express or implied warranties.
Notes: The simple portfolio consists of 70% US equity and 30% high-quality corporate bonds. The simple portfolio including cash consists of 50% US equity, 30% high-quality corporate bonds, and 20% cash. The diversified portfolio is the average portfolio for clients with more than US$1 billion in investable assets. Simple portfolio with no cash and puts assumes 200 bp yearly drag from an annually rolled 12-month, 10% out-of-the-money put option on the public equity portion of the portfolio.

Figure 5. Central Bank Balance Sheet Expansion
Source: Thomson Reuters Datastream.

Figure 6. Cumulative Wealth of Japanese Equities, Bonds, Cash, and Consumer Price Inflation
Sources: J.P. Morgan Securities, Inc., MSCI Inc., and Thomson Reuters Datastream. MSCI data provided “as is” without any express or implied warranties.
Note: Japanese bonds are represented by the J.P. Morgan Japan Government Bond Index, Japanese equities by the MSCI Japan Index, Japanese cash by the J.P. Morgan Japan Cash 3M Index, and Japanese inflation by the Japan Consumer Price Index.