Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Nokia (NOK) -- Ignore unavailable to you. Want to Upgrade?

To: Maurice Winn who wrote (1293)	12/27/1998 4:53:00 PM
From: Clarksterh	Read Replies (1) \| Respond to of 34857

To: Maurice Winn who wrote (1293)	12/28/1998 4:15:00 PM
From: Clarksterh	Read Replies (2) \| Respond to of 34857

Maurice, Tero - A post with a little more meat: But with those very low prices for GSM and cdmaOne minutes, cdmaOne doesn't really have any advantage over GSM. The capital cost of a minute is so negligible compared with the 10c per minute being charged now [usually much more than that] that the alleged capacity claims of cdmaOne over GSM are irrelevant until price competition heats up a LOT more and heads down to the 1 cent per minute range. I think your assumption that the capital costs are negligible is perhaps a little optimistic. I assume that you reached this conclusion after reading the article on the cdg website? If so keep in mind that that article was intended to be a comparison between GSM and CDMAOne, not a discussion of absolute costs. As such they didn't really explain what each of the values really means, ... . I suspect that they are talking interest expense per BS for equipment only or some such thing. The reason I say this is that when I do similar costing studies using publicly available numbers - $150K for BS equipment, 30 users per BS for a 1.25MHz channel (CDMA-2000 specs degraded for 14.4 voice - optimistic for current CDMAOne), a 5 year life for the equipment, 25% usage over 19 hrs per day (from article) - I get 3M billable minutes per year, and capital expenses of $40k per year which equates to $0.013 per minute. Sectorization probably saves some money in this area, but not much - some equipment is shared, but more equipment needs to be added and the added equipment isn't used quite as effectively used as when it was only one sector. However this does not include land costs, licensing fees, ... and these typically add up to much more than the cost of the BS (note that this is one of the places that CDMA should get a big advantage over GSM. The frequency reuse during sectorization probably saves some equipment cost/user, but it really cuts down the other costs per user. In contrast GSM does not have the same capability). Bottom line - I suspect that currently the capital costs + fees for 25% usage on a single sector per base station system is on the order of $0.04 per minute for CDMAOne and $0.05 per minute for GSM (equipment cost is a little less, but 30% fewer users per basestation so fixed costs per user go up), but that as the loading goes up and sectorization happens it will drop to $0.017 per minute for CDMAOne and GSM only drops to $0.033. (GSM's problem is that if they add six sectors they only double to triple the number of users(?) but make the system much more complex. Therefore they get to spread the fixed costs out, but the variable costs per user may even go up.) Note that even my conservative estimates for capital costs are vastly less than the capital costs at Sprint which are $30 per month per user. If you assume 300 minutes per user per month this is $0.10 per minute. If you further assume average usage north of 10% they have capital costs that are about 4 times what I would expect from my very simplisitic analysis, but are much further from the cdg article costs (Sprint capital cost = 40x cdg article). Interestingly the Sprint costs are about what I would expect for capital depreciation + site fees (land + license fees, …). The only way to make the rates in the cdg article match the Sprint costs is to assume 300 minutes per month per user and 1.25 percent average usage of the system or 1000 minutes per month per user and 4% average usage or ... . These both seem improbable, and in any case I think the numbers given in the cdg article are not what they appear to be. (I am not implying that they are wrong, just that they were meant for a different purpose.) Clark PS Note also another problem with the cdg article and its assumption of 25% usage or higher. Although 25% usage may be achievable, anything much higher is unlikely since a cell system has to be designed for peak loads. The usage in a business area drops off to negligible outside of the hours 7:00 to 19:00, and even within that time the usage outside of the commute times is probably 1/2 what it is during commute times. This loading problem is even worse for sites along commuter routes and for those rarely used sites needed just to assure complete coverage (of course the loading is probably more consistent in residential neighborhoods.) In any case I wouldn't expect too much improvement beyond 25% usage. PPS Sorry about the previous post, but discovered a mistake in my logic and didn't have enough time to fix it so I removed the post altogether. That used to result in the post disappearing after 24 hrs, but ... .