ARM Holdings (ARMH) is a semiconductor intellectual property company. It designs key functional blocks for integration into chips and licenses them out to companies who then design full chips around those blocks. Those chips are then either built by the chip designer or outsourced to a third-party semiconductor manufacturing company.

From what I understand, ARM gets paid a higher royalty percentage when customers use more cores in a design. So a chip that uses two ARM Cortex A57 cores will command a lower royalty rate than a processor that features four Cortex A57 cores and four A53 cores in what the company calls big.LITTLE configuration.

It would seem, in light of this, that the latest Qualcomm(QCOM 1.28%)processors should be a pretty big win for ARM.

More cores!
Qualcomm recently refreshed its product line-up in the low end (Snapdragon 400) and the mid-range (Snapdragon 600). In the table below, I show which ARM cores and how many of them Qualcomm put in the prior generation Snapdragon 400/600 parts, and what the newly announced ones will feature:

Old

ARM IP

New

ARM IP

Likely Impact

Snapdragon 410

4x Cortex A53

Snapdragon 415/425

8x Cortex A53

Increase in royalty from doubling CPU count

Snapdragon 610

4x Cortex A53

Snapdragon 618

2x Cortex A72 + 4x Cortex A53

Increase in royalty from increasing CPU count; added CPUs are higher value cores

Snapdragon 615

8x Cortex A53

Snapdragon 620

4x Cortex A72 + 4x Cortex A53

Same core count, but 4x A53s become 4x higher value A72s, meaning higher royalty per chip

Source: Author analysis based on Qualcomm product pages

It seems that across the board, Qualcomm is either adding more ARM IP, or it is replacing lower-value IP with higher-value IP. This should lead to a royalty rate improvement for ARM relative to prior Qualcomm chips.

A mild offset to the good news for ARM
While the increased ARM IP content inside of these new Snapdragon processors is certainly good for ARM, the potential downside is that Qualcomm's principal competition in the low-end and mid-range -- MediaTek -- often uses ARM graphics processor, and often ARM physical IP, too. I suspect Qualcomm does not use ARM physical IP and does its own physical implementations of these cores.

So, if these products lead Qualcomm to gain lots of share against MediaTek -- which also seems to be implementing lots of ARM IP (many of its parts use 8 ARM cores) -- then that could offset some of the positive effects for ARM.

That being said, the secular trend in the low-end and mid-range of the mobile market seems to be toward more cores and 64-bit cores, both of which lead to improved royalty rates per chip for ARM.

What does this mean for the business?
It is pretty well accepted that the smartphone market is growing, and given that Intel still has not been able to make a meaningful dent in the smartphone market, ARM royalty rates should grow nicely as overall smartphone unit shipments grow. That growth is juiced by what appears to be an increase in royalty rate per mobile applications processor.

Mind the valuation, please
While I like the ARM business right now (and I own the stock), I should caution more conservative investors that the shares trade for a very rich valuation. Analysts expect ARM to earn $1.42 per ADR in 2015 and $1.68 in 2016. This means the stock trades at about 30.38 times 2016 estimates.

Further note that these appear to be non-IFRS numbers (similar to non-GAAP in the U.S.), which means these numbers exclude share-based compensation. That means the "true" earnings estimate (i.e., ex-share based compensation) is lower, and the valuation is even higher.

So, while ARM is a high-quality company, and while it might be a solid stock for growth-oriented investors willing to pay high multiples for said growth, risk-averse investors may want to look elsewhere.