Will Google’s Robot Pick Good Stocks?

bokchoy888

This column is being brought to you by a robot. OK, not really, though I’ve been accused of writing like one. And something tells me my editors wouldn’t mind having a robot reporter. No breaks. No mistakes. No smart-alecky answers.

But is the pejorative “accuse” the right word in light of things that computers, robots, and artificial intelligence can now do? Perhaps in the future being called a robot—which derives from the Czech robota, meaning forced labor or slavery—won’t be so bad.

I’m thinking of last week’s fascinating demo by Alphabet (ticker: GOOGL), the parent of Google, of the Google Assistant, a virtual assistant for mobile and smart-home devices. In a phone call, the assistant made an appointment with an unsuspecting hairdresser and apparently fooled the person into thinking she was talking to a human being.

It’s an example both terrifying and exhilarating, and demonstrates that AI is passing the Turing Test. Developed by Alan Turing in 1950, the test basically says that if a computer or machine exhibits behavior indistinguishable to humans from that of a person, then it demonstrates intelligent behavior.

Now, the fear in many minds—fueled in part by Hollywood’s films about robot-engineered dystopias—is that AI, machines, and robots will steal jobs from us. A descendant of the Google Assistant might even turn into the Terminator.

Nonsense, says Joseph Quinlan, head of market and thematic strategy at Bank of America Global Wealth & Investment Management. “Rarely has rhetoric been so divorced from reality,” he says. The rhetoric is that the steady march of machines will wipe out millions of jobs in the not-too-distant future, leaving workers with less work, less income, and less of a future, he notes.

While some workers will lose their jobs, history suggests that in the long run many more jobs will be created—typically more highly skilled and better-paying jobs. In other words, nobody cries for the blacksmith now, and certainly not the many millions of car mechanics in the world.

In markets, AI has already been put to work. Robo-advisors suggest asset allocations, and many institutional investors pick stocks exclusively through algorithm-driven models.

NEWSLETTER SIGN-UP

I can handle calling the barbershop on my own, but what I’d like to know is when Google Assistant will be able to pick winning stocks. For now, machines can’t predict where markets are headed with great accuracy. Then again, neither can humans, but we keep trying.

Along with bringing flowers and, finally, warmer weather to the Northeast, the month of May has been kinder to investors. After a February to April roller-coaster ride, stocks are up 3% this month.

This mini-rally, which has returned equities to positive territory on the year, has investors hoping the market will disprove, for a third straight year, the old Wall Street saying: Sell in May and go away. In the May-October periods of 2017 and 2016, stocks were up. Call us party poopers, or worse, but some things suggest that the old proverb shouldn’t be retired. Yes, the market’s valuation, at 17 times earnings, is cheaper than its price/earnings ratio of 20 in January. Earnings growth is strong, and inflation is seemingly in check. Still, there are reasons—besides rising interest rates and oil prices—to look for continued choppy action through the fall, and perhaps to the November elections in the U.S.

First, historical data back up the adage. According to Bespoke Investment Group, $100 put into the Standard & Poor’s 500 index since 1945, but only during the November-April period, would be worth $8,100 today, excluding dividends. If the money went into the market, but only during May-October, you would have $220.

While U.S. earnings growth looks to be the best in many years—profits are expected to climb 20% this year from 2017’s level—the most recent pace of change causes disquiet. Nicholas Colas, co-founder of DataTrek Research, notes that analysts haven’t lifted their 2018 second-quarter earnings-per-share numbers yet, even after bang-up first-quarter profits. At March’s end, the second-quarter EPS and sales-growth consensus was 18.8% and 7.8%, respectively. Pretty good. At April’s end, it was 18.8% and 8.2%. What’s not to like?

Here’s the rub, according to Colas: That analysts aren’t hiking second-quarter numbers after a strong first quarter, as they normally would, implies a lack of confidence. More important, he asks, why isn’t EPS growth keeping up with revenue increases? It hints, says Colas, that costs could be rising faster than sales. Things look good, but at the margin, not quite as good as the market expects.

Moreover, while second-quarter profit estimates remain high, that is largely due to the energy, materials, and tech sectors. Everywhere else, analysts are cutting growth estimates, on average.

Another concern is the potential slowdown in economic growth. Entering 2018, the overwhelming sell-side consensus was that the global economy would remain strong, notes Dario Perkins, a managing director at TS Lombard in London. “Five months on… sentiment indicators have turned down and gross-domestic-product data generally have disappointed,” he says.

After peaking in January, about 74% of the global composite purchasing managers indexes now show decelerating growth, says Darius Dale, a managing director at Hedgeye Risk Management. About 80% of the 46 countries that Hedgeye tracks are showing or moving toward slowing growth and accelerating inflation. In general, he expects this to continue through the end of 2018. But neither Dale nor Perkins is predicting a recession.

Numbers aside, there is the potential for heightened political uncertainty in the U.S., ahead of the midterm elections. If, for example, the Republicans lose control of one or both congressional houses, what does that mean for tax reform and the easing of regulations by the Trump administration?

Let’s be clear. A bear market isn’t the worry. A return to the norm is.

Come July, Valeant Pharmaceuticals International (VRX) will change its name to Bausch Health Companies. With the stroke of a pen, the name Valeant will disappear. What won’t disappear are the residual effects on the company of a disastrous debt-fueled acquisition spree that culminated in a 97% drop in the stock to a low of $8.50, from $262.50.

If only Valeant’s fundamental challenges could be erased as easily as its name. Some investors think the change in moniker is a signal that the company has turned a corner. Its shares are up 40% since March 2, to Friday’s $20.85. Bulls like what they saw last week in Valeant’s better-than-expected first-quarter report. We were less impressed.

Valeant’s $2 billion in sales topped analysts’ forecasts slightly, and the quarter saw the first organic sales growth since 2015’s second quarter. Valeant posted a quarterly net loss of $2.7 billion, or $7.68 a share, based on generally accepted accounting principles, compared with $628 million, or $1.79 a share, of income in the year-ago quarter. Much of the red ink came from nonoperating charges, such as a $2.2 billion asset-impairment write-down. The year-earlier period included a $908 million tax benefit.

The market looks to the future, and the excitement is based partly on a non-GAAP number: Valeant raised its guidance for 2018 adjusted earnings before interest, taxes, depreciation, and amortization by $10 million, to a range of $3.15 billion to $3.3 billion. Additionally, the Bausch+Lomb/International division saw organic sales grow 2%, and branded drugs, 8%.

But then there’s Valeant’s $25.3 billion of debt. Paying it down will consume most of the company’s future income for a long time. In the first quarter, the company repaid approximately $280 million of debt.

Valeant has effectively pushed debt maturities beyond 2019, giving it some breathing room. However, that comes at a price, with new bond-coupon rates is above 9% replacing 6% debt, for instance. “That’s kicking the can down the road,” says a bearish Dimitry Khmelnitsky, an analyst with Veritas Investment Research.

Valeant’s drug pipeline might not enable it to replace revenue reduced by generic challengers and a loss of exclusivity on some products, he says. Valeant is hampered by a “fundamentally weak business” built on a different model, namely acquisitions to increase revenue, he adds.

Valeant can’t buy revenue anymore, and doesn’t have a strong enough pipeline to replace the erosion expected in products with high profit margins, Khmelnitsky says. He sees a potential $770 million in 2018-2019 sales lost from generic challenges, after nearly $500 million last year. In 2017, Valeant’s total sales were $8.7 billion.

Valeant spends only 4% to 5% of sales on research and development, versus the industry’s 15%. Nor is Valeant cheap: Enterprise value (market cap, plus net debt) is 10 times 2018 Ebitda, higher than peers’ 8.5 times average.

Valeant didn’t respond to a request for comment.

Valeant remains the subject of government probes—particularly over its former relationship with Philidor Rx Services—by the Securities and Exchange Commission and several states.

Has Valeant turned the corner? If so, future share gains could be slow in coming. But it probably wouldn’t take much bad news to pull the rug out from under the stock.

源济