Every day, Wall Street analysts upgrade some stocks, downgrade others, and "initiate coverage" on a few more. But do these analysts even know what they're talking about? Today, we're taking two high-profile Wall Street picks and putting them under the microscope...

With its shares up more than 31% over the past year, defense industry heavyweight Northrop Grumman (NOC 1.35%) is the best-performing, big, pure-play defense contractor of 2017 -- just edging out rival Raytheon (RTN) and its 30% gain. And Northrop Grumman could keep on beating Raytheon in the future...

Or at least, that's what JPMorgan thinks.

F-35 fighter jet

Lockheed Martin's F-35 fighter jet is one key to why JPMorgan likes Northrop Grumman stock. Image source: Getty Images.

Upgrading Northrop

Earlier this morning, investment megabanker JPMorgan announced it is upgrading shares of Northrop Grumman stock to overweight (and downgrading Raytheon to neutral). As TheFly.com noted today, JP thinks Raytheon is still "well positioned" in defense, but at its current valuation of 25.1 times trailing earnings, simply costs too much justify any further investment (and I agree).

In contrast, Northrop Grumman stock sells for a somewhat less egregious 22.4 times trailing earnings. What's more, according to JP, while Raytheon may be "well positioned," Northrop Grumman is "best-positioned" in defense.

What makes Northrop Grumman the "best"?

Why does JPMorgan say this? As explained in a write-up on StreetInsider.com (subscription required), Northrop Grumman is likely to enjoy a "strong year" in 2018 as the U.S. Air Force's F-35 fighter jet program ramps up production.

Now, as you're probably well aware, Lockheed Martin (LMT 0.30%) is actually the prime contractor on the F-35 program. But as part of Lockheed's team, Northrop Grumman plays a critical role in building the jet. To cite just a few of Northrop's contributions, before Lockheed Martin can assemble the F-35, it has Northrop Grumman build the airplane's center fuselage. And after assembly, Lockheed equips the F-35 with Northrop Grumman-built infrared search-and-track and fire control radars, as well as a Northrop-designed avionics suite.

In short, every time the Air Force buys a $95 million F-35 from Lockheed Martin, a good chunk of that change goes to Northrop Grumman.

The best -- and better together, too

Another thing JPMorgan likes a lot about Northrop Grumman is the company's pending merger with Orbital ATK.

When Northrop Grumman announced this $9.2 billion merger back in September, management forecast that combining the two companies would yield "annual cost savings of $150 million by 2020" and "be accretive to EPS and FCF per share in first full year." Given that over the past year, Northrop booked $2.4 billion in net profits on $25.6 billion in sales, while Orbital ATK booked $313 million in profits on $4.7 billion in sales, this implies that the combined firm will start off with annual sales of about $30.3 billion, and earn about $2.7 billion on these sales -- about an 8.9% net profit margin. All else being equal, assuming Northrop's promised cost savings arrive as planned, merging these two companies should squeeze out enough costs to lift that profit margin about 30 basis points, to 9.2%.

Does that not sound like much of an improvement? It may not be. But JPMorgan believes that management's guidance on savings from the merger is "conservative," and leaves room for "additional upside" -- meaning that post-merger, a combined Northrop Grumman could grow faster and be more profitable than promised.

Caveats and quibbles

I will not dispute that possibility -- but even so, I have to admit to having some reservations about JPMorgan's endorsement of Northrop Grumman stock today. The reason: Of the two companies entering into this merger, Northrop Grumman is by far the larger partner. And while Northrop Grumman has been growing its earnings (as calculated under GAAP) nicely these past couple years, the company's actual cash profits have more or less stagnated.

From $2.3 billion in 2012, Northrop's FCF declined to $2.1 billion in 2013, $2 billion in 2014, and $1.7 billion in 2015. After bouncing back a bit to $1.9 billion last year, FCF is on the wane again in 2017, falling to a recent low of just $1.6 billion in cash generated over the past 12 months. The company also hasn't reported FCF in excess of reported earnings since 2013, indicating that not only are Northrop's cash profits declining, but the quality of Northrop's reported earnings is weakening, even as the numbers it reports are rising.

Long story short: I'd be a whole lot more confident in Northrop Grumman deriving extra earnings from its merger with Orbital ATK if the company could show just a bit more cash to back those earnings up.