FOOL CONFERENCE CALL SYNOPSIS*
By Debora Tidwell (MF Debit)

Shiva Corporation <% if gsSubBrand = "aolsnapshot" then Response.Write("(Nasdaq: SHVA)") else Response.Write("(Nasdaq: SHVA)") end if %>
63 Third Avenue
Northwest Park
Burlington, MA 01803
(617) 270-8300
http://www.shiva.com/

UNION CITY, Ca., January 9, 1997/FOOLWIRE/ --- Shiva Corporation held a conference call after the market close on January 7th to discuss preliminary fourth quarter results. Shiva currently estimates that revenues for the fourth quarter will be approximately $48 million to $50 million. The consensus estimate for the quarter was $64 million in revenue. They anticipate earnings per share to be between $0.05 and $0.07 versus the consensus of $0.21 for the quarter. Based upon this estimate of fourth quarter results, 1996 revenues would be between $200-$202 million and earnings per share without merger expenses would be in the $0.60-$0.62 range. They would be $0.54-$0.56 with merger expenses. They will be announcing final results for the fourth quarter and FY 1996 as previously scheduled, on January 23rd.

WHAT HAPPENED. They are disappointed with these results. Overall revenue and earnings were lower than they had hoped. They have underachieved this consensus principally based on developments late in the quarter. They did not receive the Q4 product orders that they anticipated from IBM as a result of IBM's renewed focus on inventory management at year end. This was approximately a $5 million shortfall. After reviewing the IBM neoximately $7 million less than planned. Notwithstanding this situation, their book-to-bill ratio in the quarter is greater than 1 for the quarter.

UNCOMFORTABLE WITH CHANNEL INVENTORY LEVELS. While in some cases they had orders to replenish their distribution partners, they became uncomfortable with the channel inventory levels that such shipments would result in. For example, while they had promotional programs underway for LanRover in Q4, these programs did not result in the sell-through levels that they were expecting by quarter end. They felt that now was the right time to reduce channel inventory levels to try to reduce the backend loading pattern of recent quarters and to position them better for the transition that is currently underway from the fixed port market to the concentrator market. They reduced the number of weeks in the channel by approximately 5 weeks, bringing inventory levels under 13 weeks, and they are probably going to continue to bring the weeks down in Q1, particularly as it relates to the LanRover product in preparation for a new product introduction. They would like to be at 7 to 8 weeks ideally on a steady basis.

PRICE REDUCTIONS. The company was asked about price reductions during the quarter and they replied that they didn't make price reductions on the LanRover in Q4. They instituted a program that was intended to deliver an effective price reduction to the end user through a rebate program of about $1000. They did take down the classic LanRover which is the one without modems about $1000. These two moves, as they said, they had a pretty good couple of weeks at the end of the quarter did not have the anticipated effect on sell-through that they had build into their plan.

THE TRANSITION. Shiva is in the middle of a transition from the majority of their revenues coming from the stackable fixed port remote access server market to a day when the majority of their revenues will come from the higher density concentrator server market. The stackable market is represented by their LanRover product line and the concentrator market by the LanRover Access Switch line. The Access Switch shipped for the first time in the first week of June. This transition parallels that phenomenon chronicled in recent market research reports which project substantially higher growth in market size for the access concentrator market than for the fixed port market. Shiva too is experiencing this type of growth phenomena. The LanRover line has had relatively slow growth during 1996 and the LanRover Access Switch is growing at a much faster rate. In fact, they expect that the LanRover Access Switch revenue will exceed LanRover revenue sometime in the next two quarters. They think, overall, they will still be experiencing net growth for the two product categories of around 60% for 1997 even though the fixed port market industry growth rate has dropped substantially from the 80-100% rates cited by industry analysts earlier this year.

WEAKNESS NOT RELATED TO PRODUCT ISSUES. They are, in fact, in a very strong product cycle currently. Their products have received good critical response from both industry observers and many customers. They do not believe this revenue weakness is based on product issues. They continue to receive industry awards and press awards on the LanRover and the LanRover Access Switch. The Access Switch has now been purchased by approximately 400 corporate customers since it began shipping in August. Additionally, more recently they have begun to see real momentum in the telco and ISP space as evidenced by their recent announcement of wins at several leading ISPs including UltraNet, Canada Online, Surf South, and CIA Net.

COMPARISONS TO COMPETITIVE PRODUCTS. Cascade's competitive product is aimed at a very high end POP and it has good backplane characteristics and good back-end I/O characteristics for that reason. But, it also has built in some fairly high cost of entry concepts and is not a very modular product. The biggest issue in Shiva's assessment with the Cascade product compared to Shiva's Access Switch is a very weak base of software and firmware for handling hundreds of individual remote access calls which have peculiar characteristics. Shiva thinks there is everything from modem firmware to modem management firmware and software and network management to handling of PPP and the various flavors of that to all the things that have to do with authentication, authorization, accounting, and acceleration. Shiva feels they have hundres of man-years of software and firmware in that cluster of functionality and they think it is going to be awhile before Cascade can really deliver that kind of a robust system into this marketplace.

The Ascend box is a different story. Shiva thinks that Ascend obviously has a lot of in-market knowledge and a proven ability to seal costs. Their product also has a fairly high-end entry point and leaves it somewhat vulnerable. The overall configuration comparison between the TNT and a Rapport 670 (Nortel's configuration of Shiva's Access Switch) is fairly similar density using today's modem technology and fairly similar processing power. Shiva thinks what they have going for them at this point in the market is a different memory architecture, much more aggressive use of distributed processing and so far this has allowed Shiva to handle in production a lot more calls than they have heard about from Ascend installations. The main difference is in the enterprise market they come down to multi-protocol support and that includes increasingly support for tunnelling protocols. They are involved with partners in both PPTP and L2F tunnelling, support for the mainframe protocol for Token Ring, excellent IPX support, and then the system characteristics mentioned that add up to a greater ability to handle lots of calls. The strengths of the Ascend system are a far greater installed base in ISPs and, therefore, they certainly would have more experience with pure TCP-IP environments than Shiva has and with some of the peculiarities of the various ISP environments and finally they are further down the road than Shiva is in handling the GigaRouter IP switching and the backend I/O such as direct frame relay which Shiva does not yet support in their box.

Shiva has always paid a lot of attention and given a lot of respect to 3Com but have not seen them as a major player to-date. They have heard some people say that 3Com's Access Builder 5000 looks like 3Com's best offering yet, but they don't have any analysis that says that product is a player yet. In the carrier and ISP markets, Shiva's competitor is Ascend and in the enterprise concentrator market Shiva's competitor is really Cisco with some action from Bay.

CHANGE IN NORTEL RELATIONSHIP -- BACKGROUND. Shiva is going to focus more and more of their attention on 3 areas. First, they are entering into a new phase of their relationship with Nortel. They have recently deepened their relationship with Nortel with a signed memorandum of understanding. They expect to sign a definitive agreement in the first quarter that will finalize all details of this newly revised relationship. They had two obstacles to getting the kind of share they think the product and the relationship with Nortel can deliver. The first was the structural price problem they had with trying to get two manufacturing gross margins to be competitive out there against people like Ascend who are selling direct to the carrier. If Shiva is trying to get a 55% gross margin and Nortel is trying to get a 50% gross margin they end up going into the deal structurally priced uncompetitively against Ascend. Shiva wanted to change that and one of the ways to fix it is to get rid of one of the gross margins. Shiva felt it was a worthwhile effort, given the criticality of getting market share this year while the carrier market is forming and it is not yet defined who the winners are. They have some advantages with their product and the Internet throughway technology to really push this early share position.

NORTEL DEAL HIGHLIGHTS AND BENEFITS. So Shiva proposed a shift to a royalty based relationship. And there are other aspects to the deal. There are manufacturing rights payments which Shiva collects from Nortel for hardware they design and implement for Nortel. This relationship will move Shiva to a royalty-based arrangement and the funding by Nortel of the dedicated Shiva Service Provider Group and dedicated R&D efforts. This is a strategy of Shiva's that they have been pursuing for a good while. They think it will help overcome key hurdles to achieving the significant market share position in the rapidly expanding carrier access infrastructure market. For example, the movement to a royalty structure will enable Shiva to benefit from their excellent cost positioning and translate that into effective price positioning in the carrier buyer market. In addition, the dedicated service provider group with Shiva employees will deploy sales and field SE resources globally to work intimately with Nortel staff in winning and deploying carrier access infrastructure business. This will give Shiva a level of contact and control in winning this carrier business that makes them very positive about the relationship. They believe this will help improve their ability to capture a significant share of the carrier market by coupling their own Shiva personnel in the field with Nortel expertise products and historical relationships.

While they have still not made their Internet throughway announcement in a large deal with a telco, they have been awarded this business and substantial progress has been made with their telco customers. At this point, Nortel is still obligated to a non-disclosure agreement that prohibits unilateral public statements.

TARGET OTHER SEGMENTS OF THE CONCENTRATOR MARKET. They are going to increase their focus on the concentrator opportunity in the other segments which are pursuing concentrator technology, especially the enterprise market and the small to medium ISPs. They will continue to pursue with a direct sales effort in concert with VAR systems integrators and other OEM partners. What they will begin to do in the first quarter is build some focused sales, marketing, and R&D teams to address the unique needs, opportunities, and challenges of each of these markets. The low end of the concentrator business is products that deliver about a T1's worth of cost, they are often sold with the rated capacity to handle more than that, but the practical limit turns out to be 20-30 lines. Shiva thinks that will remain the bottom end because it only makes sense. Shiva will move into that part of the market with one or two T1-type of concentrators -- opening up the bottom end of the concentrator market. That makes a lot of sense for corporate customers and branch offices of corporate customers. That would be the 23-port range at the bottom end.

MAXIMIZE PROFITABILITY IN THE FIXED PORT MARKET. Third, they will continue programs that improve the position of LanRover. This is a maturing but growing market and they continue to harvest it and will maximize profitability in this market segment. There has been real action in the market at both the 8-port and 16-port level and Shiva has not had 16-port product in the market. Cisco and Bay have both entered the market with 16-port units and Shiva has seen some share lost to those companies. They are sure that not all of that comes from the port configurations, but it is clear that there are good sales in that 16-port configuration. There is also a pricing impact and that is one of the places where Shiva has seen pricing pressure. The 16-port configurations are priced higher than, but not proportionally higher than their 8-port units. They think that in the fixed port market, it doesn't make much sense to go higher than 16 ports because when you get up to 23 ports, it starts to make sense to just go to a T1 access -- doing trunking and going to the concentrators. But, in the fixed port market, a lot of the momentum will move to the higher end and the higher end is 8 to 16 ports. There's core business in the 2 and 4 port market and real small business. But a lot of that 2-4 port business is going over to the use of the software that comes with the NT server. It doesn't scale well from an enterprise requirement, but provides a fairly adequate service level for a very small port count customer. They have under development a new higher density, lower cost platform that is planned for the first half of 1997 which will improve their LanRover cost competitiveness. They intend to have a well prepared channel to receive this new product in the second quarter of 1997.

1997 FINANCIAL GOALS. They have prepared a preliminary 1997 business plan that contemplates this new Nortel relationship as well as their position in the corporate, enterprise, and ISP markets. The Nortel deal is a royalty arrangement and joint cost sharing arrangement and, as such, will have lower revenue but better operating margin percentages. Accordingly, due to the transition of the Nortel relationship to a royalty based agreement, their overall 1997 revenue is projected to be lower than current Street consensus estimates. They believe, however, that operating income will not decline. This Nortel transition will begin in the first quarter of 1997. As a result of this transition, the first quarter of 1997 revenues should be similar to their fourth quarter 1996 revenues, but then are projected to increase from this level. They said that they will have less in revenue from Nortel in Q1 versus Q4 and that it is probably safe to assume that will be in the range of $5 million to $6 million less, from Q4 to Q1.

NORTEL IMPACT ON REVENUES. They also feel that their business from Nortel will be more predictable during 1997 as a result of this more structured arrangement. They believe that the royalty and cost sharing arrangements with the Nortel relationship, while they may result in lower revenue in the first quarter, they expect the first quarter earnings should be greater than the fourth quarter 1996 earnings. They have an operating model of gross profits of 57-59% and they are not changing that model relative to gross profit on a full-year basis for 1997. Their preliminary 1997 business plan provides for 1997 earnings that approximate current Street estimates. Of course, achievement of their 1997 business plan is subject to their ability to achieve timely product development, increase channel sell-through and deployment of Nortel marketing.

WHAT THEY ARE SEEING IN THE MARKETPLACE. The demand out there is very strong. They are very close to their costs and understand their costs very well. They think they are very competitively costed in the access concentrator market right now. In their ISP installations today they are running full-up every modem in the box and that is getting noticed and people are starting to compare them on a cost per operating port basis very favorably with the competition. That is leading to some share gains and they think they will be able to maintain good margins in those businesses and they see very strong demand. They are coming from no share in the concentrator business in the May quarter of 1996 to gaining some meaningful wins. So there is a lot of ground to gain and the partnership with Nortel is allowing them to go into some carrier accounts where the scale of the wins is very substantial.

THE ENTERPRISE BUSINESS. In the enterprise business, the price environment is best understood by looking at what they see as a division in the market right now -- the small and medium business customer market is really different than the high end. This wasn't quite so true a year ago. The small/medium business customer has some alternatives today that really weren't present a year ago. There are some people doing tunnelling through the Internet using the PP/TP protocol, people who are not concerned about security very much. There are some people using the NT server software, and there are adequate products available at a very low cost from vendors like US Robotics. In that environment, price has become very significant. Shiva's thrust there has to be to get their 16-port product attractively priced into the market and they think they will have their channels very healthy to give them a good profitable return on that entry in the second quarter. At the high end of the enterprise market, they still see a very positive pricing environment and they believe they have the reference product in the Access Switch, they are having very good acceptance of that product in the corporate market. There, they think pricing continues to look very good. They have a lot of value added stuff on top of the Access Manager. So, given that they have been able to manage their costs pretty well and that the Nortel relationship is not just a royalty relationship but also includes some significant expense abatement, they are giving guidance that they feel pretty good about.

THE IBM RELATIONSHIP OUTLOOK. They haven't really contemplated a significant change in the IBM relationship as they plan for 1997. Their revenue stream with IBM, although it has not been separately disclosed during 1996, has remained relatively stable. They continue to see orders in their backlog for IBM for 1997 and they expect the relationship to probably remain at the same level it has been in the quarters prior to Q4. Therefore, they think it is a good assumption to expect revenue from IBM to remain flattish relative to 1996 at this point.

* A Fool conference call synopsis represents an effort to highlight the salient points of a conference call and should not be taken as an authoritative accounting or transcription of the entire event. Note: Statements made by a company other than historical information may constitute forward-looking statements for which the company can claim protection under the Safe Harbor Act. Please consult the company's filings with the SEC for information on risk factors which might cause actual results to differ materially from the information contained in these forward-looking statements.