Tag Archives: goog

What The Failure Of Shiller’s CAPE Shows About Stock Picking

Summary Empirical models are validated based on predictive success, not backtesting. CAPE has been above average 98% of the time since introduction in 1988, and thus never worked. CAPE has failed because earnings have secularly risen since 1992. Capital deployed on corporate buybacks and acquisitions has hit record levels during this time. Capital spent on hiring and expansion increases competition and wages. Conversely, financial engineering favors restrained hiring and improved margins. When major technology companies are compared, it is often the efficiency of capital allocation that is more important than the initial strength of their moat. Above Average is the New Average The New York Times , Marketwatch , the Wall Street Journal , and some Seeking Alpha articles all periodically warn that, according to “the historical best predictor”, stocks are currently overvalued. To be sure, SA offers a diversity of viewpoints, whereas the NYT publishes any opinion which is Robert Shiller’s. This putative best predictor is, of course, Shiller’s Cyclically Adjusted Price/Earning (CAPE) ratio, namely the current price of the S&P 500 divided by a moving 10-year earnings average, adjusted for inflation. Mr. Shiller first advocated this measure in 1988 . The idea seems so sensible it has been widely embraced, despite frequent complaints that CAPE has recently broken down. CAPE has not broken down. It has never worked. Since it was introduced, CAPE has spent about 98% of the time above average . The situation is reminiscent of Garrison Keilor’s Lake Woebegon, “where all children are above average.” We now have close to three decades of reasonably strong stock gains, despite a nearly incessant prediction that stocks are overvalued. Even CAPE’s rare dip below “average” was hardly impressive: following the Great Recession crash, CAPE in 2009 briefly suggested that stocks were 15% undervalued. Gains since then have been about 20% (annualized, not a one-time rise). Proper Empirical Model Testing But doesn’t Shiller’s model still deserve to be called the best historical predictor, given the century or so of data leading up to the 1990s? No. An empirical model is constructed using one set of data (“construction set”), and tested using a new set of data (“test set”). An empirical model should not even be announced if it does not work on the data used to construct it. Shiller’s model was announced in 1988, and constructed using S&P 500 data up to that date. Valid testing is based on subsequent years. Since introduction, it has almost continuously warned that it was not a good time to be in stocks. Yet stock gains over the last quarter century have been quite satisfactory. Yes, CAPE was particularly high before the crash in 2000, but even the ordinary trailing-twelve-month P/E was above 40. One does not need binoculars to see a barn by daylight. With quibbling exceptions, CAPE has been stuck on sell since construction. Even a stopped clock will eventually be right. CAPE is Not High Because of Irrational Investors The theory has not failed because of irrational exuberance lasting more than two decades. Jeremy Siegel has argued that CAPE should be higher than the average imputed from older data because improvements in accounting standards have upgraded the quality of earnings. Whether that is really true or not (Mr. Shiller disagrees), it is not the reason CAPE has failed. The answer – at least the proximate answer – is straightforward. Earnings since 1992 have not been cyclical at all, as the graph below shows. They have secularly increased. (click to enlarge) In addition to the secular earnings uptrend since about 1992, one should note that the decrease in earnings during the 2008-2009 Great Recession can be compared only to the early 1920s. Using CAPE today means assuming that a once-in-a-century event will happen again soon. Incidentally, CAPE is not ideally constructed. Because only aggregate earnings are considered, a company can actually negatively contribute to the valuation of the whole S&P 500. If you own 9 stocks with positive earnings, and I give you one more with negative earnings, the value of your portfolio has not declined. CAPE ought to have been constructed with positive definite components. If that mathematical term is unfamiliar, it simply means no company should negatively contribute to the value of the index. The problem is not hypothetical, given AIG’s losses reached $61.7B in a single quarter in 2008. The uptrend since the early 1990s has been quite strong, as noted in a thoughtful post from Philosophical Economics : “Over the last two decades, the S&P 500 has seen very high real EPS growth-6% annualized from 1992 until today. For perspective, the average annual real EPS growth over the prior century, from 1871 to 1992, was only 1%.” If earnings are rising by 6% a year, then predicting future earnings by a trailing 10-year average does not work . Concrete Example: CAPE Prediction vs Reality A concrete example should help demonstrate that CAPE has been high not because investors have continuously overpaid (for a quarter century!), but because CAPE has been too pessimistic about profits. The CAPE debate has been going on long enough to provide this handy example from four years ago : … CAPE was reported as 23.35 during the month of July 2011 on the Irrational Exuberance website ( irrationalexuberance.com ). Per an analysis frequently used in practice, comparing the July 2011 CAPE to its long-term average of 16.41 indicates that U.S. stocks are currently overvalued by 42.3%. In contrast, on July 22, 2011, Standard & Poor ‘ s reported a price-earnings ratio of 16.17. Using round numbers, stocks had a 100% gain over the 4 years since that warning, only about a quarter of which was P/E inflation. CAPE has almost continuously under-predicted future profits since its introduction. Why the Secular Rise in Earnings? Profit margins have surged to a record 10% of GDP, from historical values of about half that. One does not have to look too hard to discover what companies have been doing differently. When companies have excess capital, they can (1) invest in developing new products, (2) expand existing operations, or (3) buy stock, either their own or acquiring another company. In other words, they can increase competition, or engage in “financial engineering.” It is no secret that share buybacks have hit record levels, actually accounting for the majority of the total cash flow for S&P 500 companies. That is unprecedented. Mergers and acquisitions are also going briskly, with the WSJ reporting September 17 that $3.2 trillion has been spent so far this year (the number is worldwide, but the U.S. still certainly participating full heartedly in this orgy). Note that the Shiller CAPE method does capture the direct effect of share buybacks on increasing earnings. The “P” is market cap, and the share buybacks increase earnings per share, but do not change total company earnings (or, necessarily, total market cap). However, financial engineering has salutary secondary effects not captured by CAPE. Consider the case of Apple (NASDAQ: AAPL ) toward the end of the Jobs era, when Apple was sitting on $100B in cash. Steve Jobs asked Warren Buffett what should be done with the money. Mr. Buffett suggested share buybacks. This answer did not satisfy Mr. Jobs, but it has worked for Tim Cook, and Apple shareholders. Suppose that instead Apple had decided to introduce its own television (as Gene Munster incorrectly insisted), sell its own car, design its own CPU for notebooks and desktops, and perhaps even do its own fabrication of processors, instead of paying Samsung ( OTC:SSNLF ). One hundred billion is enough to do all these things at once. Each of these would have required new hiring, and building new plants. All that hiring would have tended to drive up wages. Also, the increased competition would have driven down prices, or at least had a tendency in that direction. In short, when companies buy back stock, or better yet, buy each other, instead of spending the money to increase competition, wages are kept down, and profit margins are higher. Who Benefits? While companies have spent preferentially to reduce competition (acquisitions), or by buying their own stock, rather than by hiring people to expand operations, wages have not kept up with economic growth. In fact, while corporate profits hit a record percentage, wages have increased only slowly since the end of the Great Recession. For someone hoping to be hired, or anyone wishing that another company would bid up his salary, financial engineering might not seem quite so salutary. From the viewpoint of shareholders, the recent fiscal discipline of companies such as Apple is commendable. Any specific product from any specific company might be a better idea than share buybacks. But for the market in the aggregate, less competition, lower wages, and higher profit margins have been a winning formula. Capital Allocation Some company CEOs are empire builders, others prize efficiency. Efficient allocation of capital can cause some investors’ eyes to glaze, whereas heavy spending on long shots can be inspirational. It is surprising how many people I’ve encountered, who are more likely to buy Google (NASDAQ: GOOG ) (NASDAQ: GOOGL ) because it is working on driverless cars, investigating quantum computing, and frittering away money needlessly indulging the founder’s whims. Conversely, Wall Street money managers take capital allocation quite seriously. The enormous surge Google had in the hours after its July earnings report had little to do with the actual mediocre results and a great deal to do with hints that the recently hired CFO, Ruth Porat, was going to bring much-needed efficient use of capital to the company. Anticipating a Counter Example Microsoft (NASDAQ: MSFT ) pays dividends and buys back shares. Amazon (NASDAQ: AMZN ) spends everything it earns from its retail operations to compete in new areas. It also has been hiring robustly to do so. Doesn’t this show that financial engineering doesn’t work? Hardly. MSFT practiced laughably poor capital allocation for at least two decades. Say the words “serial overpayer” to a market aficionado and she will likely take you to mean MSFT. MSFT also tried to emulate the great laboratories of the past (such as Bell Labs of the old AT&T (NYSE: T ), or what Xerox (NYSE: XRX ) had), hiring notable academics for long-range research. And MSFT has spent tens of billions trying in vain to compete with Google in search and Apple in phones. Amazon gave the last one a go as well. But when the Fire Phone failed, it didn’t spend another $7.2B to buy a fading phone designer. Amazon just laid off the associated workers from its Lab 126. The company is not worried about this quarter’s numbers, but the company is nonetheless very results-oriented. Bezos was a hedge fund manager before founding Amazon, and his keen interest in careful capital allocation manifests in many ways, including not overspending on employee benefits, and the practicality of the projects Amazon attempts. Admittedly, a stock buyback is rarely the absolute best possible use of money. It is, however, typically better than the empire building most CEOs attempt. Summary Part 1: So, Is The Market Overvalued? One generally can do well by simply looking at ordinary (meaning TTM) P/E, and whether that has been rising or falling. The current P/E is 20 and, alas, trailing-twelve-month earnings have been falling for about a year. Forward analyst estimates are meaningless, as documented in excruciating detail in Burton Markiel’s A Random Walk Down Wall Street . So, yes, stocks are overvalued, but by about 15-20%, versus the 40% plus suggested by the current CAPE (= 25). Believing in CAPE requires believing a once-in-a-century profit recession is imminent, and that corporations are soon going to abjure financial engineering and start more aggressive expansions – plausible, but hardly certain. Summary Part 2: Capital Allocation In The Aggregate For the market as a whole, the current large portion of corporate free cash flow spent on share buybacks and acquisitions has restrained hiring and thus wage pressure, while reducing competition. This has steadily improved net profit margins, and raised corporate profits to a record 10% of GDP. This phenomenon largely explains why profits have not been cyclic, but secularly rising. Summary Part 3: Capital Allocation In Stock Picking For the very long-term investor seeking to exploit the tax advantages of unrealized gains, capital allocation is crucial. Indeed, because of compounding, capital allocation will eventually win out. If a company fritters away its earnings, even a business as great as Microsoft’s can struggle to provide adequate returns. Conversely, if a company returns money to shareholders, and carefully monitors whether its projects are producing worthwhile results, long-term performance will be superior. Disclosure: I am/we are long AMZN. (More…) I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Tech Titans Are The Princes Of Disruptive Competition, Which To Buy?

Summary This article compares Facebook, Amazon, Apple, Netflix, Google, and Tesla, not in their competitive market environments, but in the investment market competition for higher coming stock prices. As we are wont to do, the comparisons are through the eyes of Market-Makers [MMs], protecting themselves as they serve the block-size transaction orders of their big-$ clients. But those views are importantly conditioned by big-$ clients order-flows, the ones with the money muscle to move markets. Opportunity can be created – their collective competitive motto These companies all seek to build their kingdoms by upsetting the established order in their spheres of influence. Information, retail, entertainment, transportation all have in common huge size of markets, with innovations of varied types the principal tool of disruption. Wielded by able, controversial, charismatic leaders. But equity investment markets provide a common denominator for comparisons among even the most diverse of subjects. And behavioral analysis of the players in this very serious game ensures a look at what investors actually expect in coming days weeks and months, (not what others may want you to think) in the scorecard common to all: stock price changes. Price changes initiated and supported by players with the money muscle to make things happen. Price changes to enrich – or wound – active investors who are sensitive to the value and power of the investment resource of time. The role of self-protection Inside the ropes of the investing ring, the rule has to be: Protect yourself at all times. So MMs, when exposing their capital to risk in balancing buyers with sellers in million-dollar-plus trades, buy protection to ensure the continued presence of that capital for employment in the hundreds of like deals yet to occur each day. “All” that takes, is finding some other player to take on the perceived risk – for a price. The price of the protection, and the structure of the hedging deal tells what the players think is reasonably possible to happen to the subject issue’s price during the days, weeks, and months it takes to remove the capital from risk and to unwind the contract commitments of the options, futures or swaps performing the risk transfer. This is analysis of the behavior of experienced, qualified professional experts, doing what is right, so that we can borrow his judgment to help us in our goals. Other behavioral analysis tends to focus on the human errors that folks make, perhaps so that they may be victimized. Such efforts have so far failed to identify any meaningful rewards from that approach. In each subject of this analysis there will be clear evidence of the frequency and extent of the investment rates of return previously achieved subsequent to prior forecasts like those of today. No guarantees, just the perspective of what previously has been accomplished. The basic Risk vs. Reward tradeoff We start by comparing what differences presently exist between the FAANGT stocks upside price change forecasts, and the worst-case price drawdowns in the 3 months subsequent to prior forecasts like those of today. Figure 1 (used with permission) Pictured are Facebook (NASDAQ: FB ), Amazon (NASDAQ: AMZN ), Apple (NASDAQ: AAPL ), Netflix (NASDAQ: NFLX ), Google (NASDAQ: GOOG ) (NASDAQ: GOOGL ), and Tesla Motors (NASDAQ: TSLA ). In general, these stocks present a fairly uniform tradeoff of expected returns from coming price changes against the worst experiences of what has happened in the past, following MM forecasts like those of this day. The one modest divergence is that of AAPL [2], where downside experiences have been less extreme than all others, especially vs. FB [4] and GOOG [5] which offer slightly less optimistic upsides. The biggest return payoff prospect, NFLX [1] carries the highest risk exposure prior experiences. But while this tradeoff is paramount for most investors, there are times and circumstances that raise other considerations in importance. Thoughtful investors usually have many of these other dimensions in mind at all times, with varying degrees of importance. The table in Figure 2 lays several of them out in an orderly comparison. Figure 2 (click to enlarge) What has been pictured in Figure 1 was taken from columns (5) and (6) of Figure 2. In turn, (5) is a calculation of (2) divided by (4), minus 1, expressed as a percent. Column (6) is an average of the largest negative experiences of the first sub-column of (12). All of (12) tells how many times in the past 5 years’ market days of forecasts there has been a forecast like this day’s. You may note that FB has only been around since its more recent IPO of some 3+ years ago, while most of the other “old-timers” have a full 1261 days of forecasts. TSLA is but a couple of 21-day months short of that mark. Importantly, Reward~Risk maps like Figure 1 and tables like Figure 2 are creatures of the date those snapshots were taken, not lasting comparisons of the “goodness” of these stocks versus one another over long periods of time. The forecasts involved here are implied from the “crowd-source” judgments and actions of buyers and sellers of price change protection caused by market makers seeking to fill volume (block) transaction orders by “institutional” clients managing Billion-$ portfolios who must operate on a scale that “regular-way” market transactions cannot accommodate. Those transactions normally involve the MMs having to put at risk amounts of firm capital in order to bring to balance buyers and sellers of the stock involved. Hedging of those capital risk exposures involves negotiations between the MM and sellers of such protection via derivative securities contracts involving options, futures, or swaps. The prices paid for protection and the structure of the deal tell how far the subject’s price may be likely to travel, as seen by all parties involved. That includes the investing organization initiating the trade order, since they wind up paying the cost of the hedge, as a part of the price of market liquidity arranged by the MM. It turns out that all 3 parties are well-informed, experienced, “consenting adults”, whose actions provide information that is generally not recognized. Such traffic goes on day after day, causing changes in the relative attractiveness of stocks to one another. But the implied price range forecasts of columns (2) and (3) are limited in their time scope to the life of the derivatives contracts used in the hedge deals. Those are usually kept as brief as possible, out of cost considerations. It turns out, based on decades of daily forecast experiences, that their reliability and usefulness diminishes markedly out beyond 6 months. Within that horizon, a more critical limit of 3 months turns out to be a useful boundary. Using that limit we look to see what bad things – like price drawdowns from (4) – have happened following prior like forecasts. That average of worst experiences is in (6). Column (8) tells what percentage of the forecasts have had prices recover from (6) to be profitable in reaching (2), or by the time 3 months later has arrived. At either of those points, the net of gains minus losses is presented in (9), the time taken in (10), and the CAGR in (11). Other measures of comparative interest are the proportion of (5) to (9) in (13), and the relation of (5) to (6) in (14). A figure-of-merit is calculated in (15) using the odds of (8) and its complement to weight (5) and (6), recognizing the frequency emphasis provided in (12). This Figure 2 table is ranked by (15). So, what to do with this data? Right now, it can be used to make comparisons between the probable coming near-term price movements of the six stocks. The investor may want to embed in his considerations the personal preferences he may have as to the need for price gain in the face of potential price loss, and the implications of the time horizon involved. For example, in 80 past days FB has had knowledgeable and experienced market professionals making good-sized bets that a +10% gain is likely to occur, and in 65 of those times a profit has occurred. Net of the other 15 times, a 7% gain was had on all 80 bets. That has happened in 10% of FB’s entire market existence, so it’s not a rare occurrence. On the other hand, NFLX could indeed be up by 18% within 3 months, and smart money is willing to pay to avoid being hurt that way. But in the 135 prior times his kind of forecast has appeared, the investor has seen about 1/6th of his investment disappear at least temporarily, and at least 3 times out of every ten, some of it permanently. All of that to earn a slightly smaller net payoff in about the same period of capital commitment to get a CAGR like FB’s. Then TSLA appears to offer a prospect of over +15% gain in the same period, with odds similar to NFLX and drawdown exposures of -12% instead of -16%. But its average net payoffs, a little better than NFLX, took measurably longer to be achieved, so its CAGR is markedly less. Perhaps more important is a comparison with what else is available. The blue summary lines at the bottom of Figure 2 tell what a market-tracking-proxy, SPY, currently offers, how that compares with the population average of 2500+ equity alternatives, and the best-ranked 20 of that population. SPY currently reflects the concerns of many market gloom-&-doomers with a not very credible upside possibility of under +9%, while having delivered less than +3% under similar forecast circumstances in over a year’s worth of experiences. Only GOOG presents as poor a CAGR history (from today’s forecast) as does SPY. The whole population provides some eye-catching +15% or better possibilities, but is dramatically short on deliverance, with only 6 out of every 10 profitable, and achieved gains only one fifth of the promises. The history of the 20 best-ranked stocks on the other hand is of interest. No guarantees that it will be repeated now, but in over 10% of the forecast opportunities of this group of stocks, they have delivered on price gain return prospects of +11%. With drawdown experiences of only half of the gain prospects, they have recovered in 7 out of every 8 cases to deliver net gains at an annual rate of better than 100%. On our figure-of-merit scorecard, they have collectively ranked some three times better than FB, the best of the 6 competitively disruptive stocks discussed above. Conclusion This appears to be a more opportune time to buy FB and AAPL than SPY or the other identified alternative stock candidates. Tomorrow is likely to be some different. But a good deal of what goes into choosing between investment commitments depends beyond simple measures, like P/E ratios, or ephemeral measures like what this charismatic CEO foresees. What really counts is what the investor community believes is possible for the stock’s price in times to come. And to what degree those beliefs have been borne out in the harsh reality of the marketplace. Comparison is the essential tool of the valuator. History may help in providing perspective. But equity investors need to be mindful that uncertainty is always present because investing involves the future. So some guessing about the odds of success (however it is defined) is always required. Help in that effort may be useful. Or not. What are the odds? Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. (More…) I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Goldman Sachs Issues To-Do List For Google’s New CFO

Goldman Sachs says Google’s (GOOG) new CFO should take the bull by the horns by increasing transparency on YouTube and other parts of the business and making better use of the search leader’s $65 billion cash balance, perhaps with a stock buyback. Sustaining revenue growth will also be a challenge as the core search business matures and amid growing rivalries with Apple (AAPL), Microsoft (MSFT), Facebook (FB) and others, says Goldman Sachs analyst