Source: Micron.

Micron Technology (NASDAQ:MU) is on a roll. The memory-chip maker's stock has gained more than 470% over the last two years, 62% in 2014 alone. And analysts believe Micron might have even more room to run. Piper Jaffray just reiterated its "strong buy" call on the stock, raising its one-year price target from $43 to $44. Hitting that goal would be another 26% gain from the stock's current price.

But, as Kenny Rogers would say, you've got to know when to hold 'em, know when to fold 'em, know when to walk away, and know when to run.

I'm a Micron shareholder. And while it's nice to see my investment thesis playing out exactly as planned, to the point that Jaffray's analysts list the company strengths I saw as growth drivers for the next leg of the journey, Micron is not one of those stocks I'd buy once and hold forever.

How much further can Micron fly on the current tank of gas, and what should I do with my personal holdings?

Piper Jaffray's analysis
First, let's look at Piper Jaffray analyst Ruben Roy's target-price upgrade.

In a long wave of bankruptcies and buyouts, the memory-chip market has largely boiled down to three important players. Roy said a select group led by Samsung, Micron, and Hynix add up to a sweltering 95% share of the DRAM chip market. It's a mature industry now, and acting like one.

The three top suppliers have shown discipline in their manufacturing strategies, holding back on factory construction and letting the supply and demand balance support higher market prices. On the demand side, PC memory is being replaced by new growth drivers, as cloud servers, mobile devices, and machinery built for the Internet of Things all require memory chips.

Those two trends explain why DRAM prices are high and stable -- a welcome change of pace from the relentless memory industry price wars of the previous decade.

That's the random access memory market, where larger installations help computers performing intensive number-crunching. In the NAND chip sector, which focuses on longer-term storage such as solid-state drives and mobile data storage, the story is much the same.

Here, Roy expanded his dominating cabal to five suppliers by adding SanDisk (UNKNOWN:SNDK.DL) and Toshiba. These five companies control 96% of the global NAND market.

This sector isn't quite as stable as the DRAM market, but it's getting there.

Did you see these returns coming?
I pulled the data and discussion above from Roy's fresh analyst note and current memory market indicators. But all of it sounds like an echo from 2013, when I laid out the reasons I bought Micron shares for my retirement account.

My investment thesis is playing out exactly as I envisioned. Micron's buyout of Elpida changed the rules of the memory pricing game in the way I thought it would, and now the market has stabilized at a highly profitable level. My Micron position has gained 345% in 42 months, which works out to an average annual return of 53%.

Once again, it's time to make a decision. Should I cash out my extremely profitable Micron stake, or should I let this winner run even further?

In earlier examinations of my Micron position, I said the company deserves a price-to-earnings ratio somewhere near 20 times trailing earnings. That's a fairly industry-standard valuation. Micron used to trade at lower prices simply because its products were subject to severe price wars. That placed the company firmly in the category of offering commodity products, which limits long-term profit margins and leads investors to pay lower prices for the stock. But the pricing power that came with Micron's Elpida deal was well on its way to erasing that commodity discount, so the stock should be valued in line with its closest industry peers. 

This past summer, you could buy Micron stock for about 10 times trailing earnings. Today, the price has climbed to nearly 15 times earnings. That's not far from the 17 times earnings P/E ratio of semiconductor giants Intel (NASDAQ:INTC) and Qualcomm (NASDAQ:QCOM).

Micron can reach Roy's $44 price target without expanding its P/E ratio, since he expects earnings to grow roughly in line with his estimated share price gains. So at least one pro thinks the stock has already reached its proper price-to-earnings multiple.

Moreover, the adjusted earnings Roy's estimates, and indeed most analyst estimates in general, are based on are subject to all sorts of financial shenanigans. Since it is the most commonly watched shorthand for profitability and financial health, not to mention the basis for many executive bonus programs, plenty of companies do whatever they can to meet Wall Street projections. On the other hand, high earnings also lead to large tax payments, so management has a duty to keep those expectations low. The tension between the incompatible goals of good-looking income and thrifty tax accounting can inspire management to use many kinds of strange, incomprehensible, or downright misleading accounting tricks. 

These concerns don't apply to cash flow results, which measure the actual money flowing into and out of the company's balance sheet. If we replace Micron's share price with its enterprise value, we'll also capture the value of its strong assets and low debt levels.

From the perspective of enterprise value divided by trailing free cash flow, Micron suddenly looks absolutely comfortable at its current levels.

Micron is worth 13.3 times trailing cash flow, which is in step with the industry giants I'm using as a model. Over the last three years, the EV to FCF ratios for both Qualcomm and Intel have bounced between 12 and 20, typically clustering near 14. Right now, Intel is more richly valued at 18 times cash flow, but Qualcomm sits just above Micron at 13.8.

INTC EV to Free Cash Flow (TTM) Chart

INTC EV to Free Cash Flow (TTM) data by YCharts.

Now, the market doesn't always get every valuation right. Micron is a case study for this fact, given its spectacular share price rise while a dead-obvious investment thesis simply played out exactly as planned. If Mr. Market truly owned a perfect crystal ball, the stock would have soared when the Elpida deal was approved, and then only adjusted for normal business growth after that.

But the same market sets the trading prices for every stock, warts and all. So it's useful to look back at longer-term trends, with the understanding that nothing is set in stone. And that's why I'm comfortable with Micron's valuation multiples today, seemingly based on proper investor respect for the company's cash flow.

The final verdict
In the final analysis, I'm looking at a properly valued stock with a rock-solid fundamental business. There won't be any quick doubles or triples for Micron investors -- the buyout-powered catalyst for that has already done its job.

Against that backdrop, I'd expect Micron shares to rise more or less in lockstep with the growth of its core business. Market prices for Micron's memory chips have stabilized, in large part thanks to the company's own supply controlling actions. And we can play buzzword bingo over the growth drivers on the demand side.

So Micron isn't the no-brainer extreme growth stock it once was, but it remains poised to grow faster than your favorite market index over the next several years.

Never count your money while you're sittin' at the table, Rogers said. Fair enough, but Micron has become an oversized slice of my investment portfolio, and I think it makes sense to lock in some of those fantastic profits. So I'll trim my Micron position somewhat in the near future, but I have no intention of closing out this investment entirely.

There's still too much value in Micron to make me walk away, and I wouldn't even consider running.