Loading...
HomeMy WebLinkAboutCOW 2008-09-22 Item 4C - Discussion - Tukwila Village Selection of Developer -j CO UNCIL AGENDA S I'NOPSIS .VILA Initials ITEM No. S Q VOX 1 I P a 3 9 1 �Meetinp, Date 1 Prepared by Mayor's Council review 1 i S s3 1 08/18/08 1 DCS I 1_ 1 4, 0--, 1.. 1 08/25/08 1 DCS 1 1 1 1 c+- rsos -S_�- 09/08/08 DCS 1 09/22/08 DCS 1 AZ 117 1 SP B' ITEM INFORMATION I CAS NUMBER: 08-109 I ORIGINAL AGENDA DATE: AUGUST 18, 2008 AGENDA I1'EM TITLE Tukwila Village: Selection of developer C. \1'IGoRY Discussion Motion Resolution 1 Ordinance Bid Award Public Heart: nb Other lltg Date 09/22/08 Mfg Date 09/22/08 Mtg Date Mtg Date Mtg Date lltg Date Mtg Date I SPONSOR Council IVlayor Adrn Svcs DCD I I Finance Fire legal P6R Police [1 PW SPONSOR'S Two developers have submitted proposals to develop Tukwila Village. SUMMI;1RY At this meeting, Council may discuss the proposals and select a developer. RI.1'II.AVI {D BY COW Mtg. CA &P Cmte F &S Cmte [1 Transportation Cmte Utilities Cmte fI Arts Comm. Parks Comm. Planning Comm. DA 1'E: RECOMMENDATIONS: SPONSOR /ADMIN. Discussion /Select one developer COMMITTEE. COST IMPACT FUND SOURCE ExPI•,NDITURI!: REQUIRED AMOUNT BUDGETED APPROPRIATION REQUIRED $0 $0 $0 Fund Source: Continents: 1 MTG. DATE RECORD OF COUNCIL ACTION 08/18/08 Developer proposals were presented to the Council 08/25/08 I Discussion and questions and answers; forward to 9/8/08 Council Meeting 09/08/08 Discussion and questions and answers; forward to 9/22/08 Council Meeting MTG. DATE I ATTACHMENTS 08/18/08 Informational memo providing a basic statistical comparison of the proposals dated 8/12/08 Informational memo comparing statistics for proposed housing dated 8/14/08 Informational memo comparing developer experience dated 8/14/08 08/25/08 Informational memo evaluating the proposals to the City's criteria dated 8/21/08 09/08/08 Informational memo providing a revised evaluation of the proposals dated 9/5/08 09/22/08 1 Informational memo with Fuller /Sears /Ravenhurst report dated 9/15/08 (Please bring the 2 proposals from the developers distributed in July) q s City of Tukwila Jim Haggerton, Mayor o 3.. j Office of the Ma iii O f Mayor Y N a e 6200 Southcenter Boulevard 1 r Tukwila, WA 98188 190$ www.ci.tukwila.wa.us Date: September 15, 2008 To: City Council Copy: Mayor City Administrator Council Analyst From: Economic Developm mini for Re: Tukwila Village Fuller /Sears/Ravenhurst Report The City contracted with Fuller /Sears Architects and Ravenhurst Development to review the two developers' proposals for Tukwila Village. The main purpose of the review was to evaluate the project financials with an emphasis on development risks and design elements that could affect the project financials. Fuller /Sears Architects is a full- service architectural firm in Seattle that was already very familiar with Tukwila Village. Fuller /Sears performed the architectural work for the Sabey proposals in 2002 and the Metrovation (Ron Sher) concept in 2007. Bill Fuller, one of the principals in the firm, provided the architectural oversight for this evaluation. Ravenhurst Development is a commercial development and real estate consulting firm based in Seattle. Ravenhurst is currently partnering with TRF Pacific LLC on the $300 million, mixed -use Dearborn Street development to be located on the 10 acre Goodwill site in Seattle. Darrell Vange is the owner and president of Ravenhurst and performed the financial evaluation. The report provides a number of conclusions. The main conclusion is that neither project appears to offer a high enough return on cost to be economically feasible, although they are close (see page 13). However, the authors note that their assumptions may have been too conservative. Using more optimistic assumptions would tip the equations so the projects would be financially feasible. The report also makes a number of comparisons between the two proposals based on various assumptions (e.g. assuming the same apartment rents or different retail vacancy rates). After these comparisons, the authors conclude "the choice of developers cannot and should not be made based on a review of the financial models." I have attached a copy of the report, including tables that summarize the four different financial statements. They are: (1) Legacy's proposal "base model" (2) Tarragon's proposal "base model" (3) Tarragon's proposal "Apples to Apples Because Tarragon's proposed apartment rents of $1.35 per square foot seemed unrealistically low, this version uses Legacy's rent, apartment sizes, and estimated land payment to the City to make more of an "apples to apples" comparison. (4) Tarragon's proposal "Apples to Apples with Retail Vacancy Because the authors do not believe the market will support as much retail as is proposed by Tarragon, this version uses the "apples to apples" comparison but also includes a higher vacancy rate on the retail. Although it is disappointing to me that the financial comparisons are not able to offer a more clear picture of which proposal is more feasible and estimates a higher value for the land, this is not surprising. There are three main reasons why it is challenging to quantify, with a high degree of certainty, the differences in the feasibility and estimated land value between the two proposals: (1) "Residual effect Since the land value is a residual result of the project revenue and costs and a small portion (approximately 5 of the total project cost, small changes in revenue and cost assumptions can cause large fluctuation in land value. (2) Conceptual stage: It is always difficult to predict a specific land value at this early stage when the design is so conceptual because there are many unknowns that affect the development. For example, the final design may have significantly different amounts and configuration of parking spaces. Without getting detailed estimates from contractors or professional cost estimators, it is difficult to predict the difference in cost between the Legacy approach as compared to the Tarragon approach. (3) Apples and Oranges: The proposals are based on very different assumptions that will probably become much more similar as the project evolves which means we should not draw too strong of conclusions from the different assumptions. For example, Legacy's proposal assumes an average rent of $1.65 per square foot yet Tarragon's assumes $1.35 per square foot. The proposals may reflect different product quality and mix which can explain some of the difference in expected rent but in the final design, the product types and mix will probably be fairly similar and so the rents would be, too. In my opinion, the Fuller /Sears/Ravenhurst report offers many good insights and suggestions. Some of the suggestions and risks can be addressed during negotiations. In terms of selecting one of the developers, the conclusions I draw from the report are: (1) Both developers are very qualified and can do this project. (2) Neither proposal is a "sure thing The projected costs and revenues show the projects are not quite fmancially feasible, but with more optimistic assumptions, it is possible the projects are feasible. (3) The Legacy proposal provides more detail and the design appears to be thought through more (e.g. the grade difference between the outdoor plaza and comer retail store in the Tarragon proposal will be challenging to remedy) so there are fewer "unknowns." Tarragon's proposal will probably evolve to be a bit more similar to Legacy's (e.g. how parking is handled, height and footprint of the apartment buildings). (4) The Tarragon proposal appears to be highly dependent on getting a 15,000 square foot drug store on the corner. This may be very possible and would support the fmancials, however, the question is whether the City would want this type of tenant on the corner. (5) The most significant unknown risk is the amount of demand for retail. This risk is higher with the Tarragon proposal since it assumes more retail. If Tarragon can get the retail at the rents they propose, it improves the financial feasibility of the project. (6) The second most significant unknown risk is how much rent people will pay for the apartments. This risk is generally the same for both proposals since they are likely to seek the same demographic of residential tenant. (Note: This comment does not take into account the difference between senior housing, ownership townhouses, and affordable apai Intents.) (7) The report also mentions that in Tarragon's proposal the parking for phase 1 appears to be too low. Parking is very expensive and so if more parking is needed, this could be a very significant cost and thus reduction in the price we would receive for the land. If you would like a copy of the entire project budget and income statement proformas developed for the Fuller /Sears/Ravenhurst report, please let me know. Please contact me at 206 433 -1832 or dspeck @ci.tukwila.wa.us if you have any questions. Tukwila Village Evaluation of Legacy and Tarragon Proposals. 1. Introduction Fuller /Sears Architects and Ravenhurst Development, Inc. have been asked by the City of Tukwila to assist in the evaluation of the two responsive proposals to the Tukwila Village RFP dated July 11, 2008. Two firms, Legacy Partners and Tarragon, responded to the RFP, and this summary memorandum contains an evaluation of the design concepts, the economic implications, the development risks and an assessment of the capabilities of the two proposals. This information is intended to supplement and not replace the memoranda and comparison materials prepared by Derek Speck for the Mayor and City Council. Fuller /Sears specializes in the design of urban mixed -use projects similar to the program proposed here, and has had a long involvement with the site, assisting Sabey in their initial planning for the site. Fuller /Sears also worked with Metrovation on a second development plan in early 2008. Ravenhurst Development is a private commercial development firm, currently developing a major mixed -use project on the Seattle Goodwill site in downtown Seattle. Ravenhurst often provides development perspective for municipalities considering or engaged in "town center" development. Each response to the Tukwila Village RFP included a conceptual design with site and building plans. While these were done by very capable architects, they are at this point just concepts, and very few of the working details have been resolved. They illustrate the mix of uses and the relative location and size of the various parts, but should not be taken very literally. Whichever developer is selected, you should anticipate that the final project may look very different than the designs presented today. Of all the elements we will consider in our review, the "look" of the project is not one of them. Both proposals include some limited financial information, but the detail provided differs greatly between Legacy and Tarragon. Rather than prejudice either proposal, we prepared complete financial projections for each project from scratch, relying as little as possible on the cost and revenue projections provided. Like the design concepts, the financial information provided is simply too conceptual to use for evaluating the projects. FULLER =SEARS Tukwila Village Developer Evaluation RAVENHURST ARCHITECTS September 12, 2008 DEVELOPMENT, INC. 1 2. Report Summary Both proposals are complex and responsible programs from experienced developers. Either one would be very challenging, and neither is likely to be executed exactly in the design and configurations as shown on the conceptual plans. The Legacy program is more thought out, is more pragmatic, and has a higher likelihood of execution. The Tarragon program is more exciting, is far more aggressive, and will probably change significantly before it gets to final design, especially in the parking and residential plans. If measured by what the city leaders want, we would guess the Tarragon plan is more appealing. If measured by what the market will support, the Legacy plan may be closer to the mark. By selecting a more visionary course, there is added risk that all the effort will be for naught, and in three years time there is still no development activity on the site. There is also the very good chance that the exciting plan gets whittled down by market realities, and ends up being very similar to the program you did not select. We suggest the selection should not be made purely on the emotive impact of the conceptual renderings. We also suggest the selection not be made on the financial information, as the numbers provided by the developers are incomplete, and the marketplace will bring the figures closer together. Our assessment is that neither program is likely feasible in today's financial markets, and further concessions and/or support will be necessary to close the deal. Both developers are qualified. Both will refine their plans as the project moves forward. Both have to live in the same retail marketplace and the same financial community. We suggest you assess the developers and not the plans. Consider their level of interest, the commitment and effort they have invested to date, the rapport and level of trust they have built in this process. If you have not spent enough time with them to make this kind of assessment, invest the time. And even though you may not be real estate experts, you live in this community; you have a gut level feel for what is realistically possible for the site. Select the developer you feel can deliver what they promise, both in terms of their character and their vision for the site. Don't make it easy by trusting their vision; make them convince you their program is possible. FULLER SEARS Tukwila Village Developer Evaluation AR CHITECTS Se tember 12 2008 RAYENHURST p DEVELOPMEAT INC 2 3. Design Character Overview Legacy: The Legacy development program appears to respond to all of the programmatic requirements of the RFP. It includes the library in a prominent position at the TIB edge of the site; it includes the police resource center at the corner of TIB and 144 in a highly visible location. The program includes a 3400 sf community room as a fully enclosed and built -out space, with an 8000 sf plaza immediately outside the community room and the library. The retail program, at 10,000sf in phase 1, is small, but it has good visibility from TIB, with adequate surface parking adjacent to it. The residential program is divided into three distinct parts: there is a senior housing building on the north end of the site, to be developed independently by LIHI, 144 units of market -rate housing overlooking the central plaza developed in three distinct buildings, and twenty -five for -sale townhouse units on half of the site south of 144 also developed by LIHI. A second market -rate apartment building is planned for south of 144 in phase 2. The overall site plan acknowledges the highway oriented character of the surrounding neighborhood, and provides a more urban, more -dense design that complements the neighborhood without pushing the limits of urban design. The most significant visual impact is the building height, as the project uses five stories of housing over one story of commercial base. The program is driven by the residential, which makes some sense, since Legacy is primarily an urban residential developer. The residential elements are efficient and pragmatic, and the inclusion of LIHI adds specialized affordable housing expertise and reduces development risk. Tarragon: The Tarragon program also responds to the programmatic requirements of the RFP, and reflects the strengths of the Tarragon organization. Their plan has a much larger retail component, witi l a higher emphasis on place making. The key featur of the plan FULLER SEARS Tukwila Village Developer Evaluation ARCHITECTS September 12, 2008 RAVENHURST DEVELOPMENT INC 3 is a large central plaza of 20,000sf, with a 1000 sf community pavilion. The plaza is ringed by parking, and surrounded on all four sides by the mixed -use buildings. The plan includes 40,000 sf of retail in the first phase, with a large retail building on the TIB frontage and at the leading corner of the project. While the residential unit count is similar, the character is very different, as the buildings have three stories (in Phase 1) or four stories (in Phase 2) of residential over a one -story commercial base, and the above -grade structured parking is distributed around the site rather than consolidated underground. The program is more internally- organized than the Legacy design, and has less visibility into the site from TIB. It may be a stronger project design, but it does not relate to the surrounding commercial neighborhood as well. Emotionally and from a presentation standpoint, the Tarragon proposal is more attractive. The rendering style is warmer, the program is more people- oriented and the amenities are more obvious. The question is whether the individual elements function as well as they need to, and whether there can be enough activities and people to fill the large public spaces. Design Details Legacy Parking strip: the Legacy plan features a parking strip along the TIB frontage, which pulls the public space back from the busy street and acts as a buffer. This is more in character with the highway- oriented development forms of the surrounding properties, and the commercial areas across the street. We consider this an appropriate response to the neighborhood conditions. 5 over 1 vs 3 over 1: the Legacy residential program features five story buildings in easily phased blocks. This is a more efficient building type than a three -story or a four -story project, with lower construction costs and shorter internal corridors. Since we expect that rent levels will be a critical hurdle in the project, building efficiency will be important. Community room: the Legacy community room is a 3400 sf multi -use space that is well located next to the library and facing the plaza. It has the potential to be a very usable community resource. At the same time, it is an expensive feature and a very large space, so it should be constructed only if there is a clear program of uses and a means to pay for its costs. FULLER==S ARS Tukwila Village Developer Evaluation ARCHITECTS September 12, 2008 RAVEN L HURST 4 Plaza utility: the Legacy plaza is much smaller than the Tarragon plaza (8000sf vs. 20,000sf), and it is directly adjacent to the library and community room, which allows for easy activation. An 8000sf plaza is large enough for a 20 -stall farmers' market or a gathering of 1000 people standing. The city should determine how they expect this space to be used, and make sure it is designed to an appropriate size, not too large or too small. Three housing types: We consider the Legacy plan, with the affordable housing elements developed by LIHI, to be an appropriate solution to the city's desire to include affordable housing in the mix. Neither Legacy nor Tarragon have specific expertise in senior or affordable housing of this type, and LIHI is an expert in this type of development. Housing density will be a critical factor in the success of the project, and providing diverse product types and a range of income levels increases the likelihood the project will be fully built -out. Grade treatment: the site slopes away from TIB, and is approximately 10 feet lower than TIB on its eastern edge. It is preferable for the plaza and the internals of the site to be at grade or slightly higher than the street edge, and the Legacy grading plan accomplishes this. It is not clear that the edges along 144 have been resolved, but the plans are at a very conceptual stage. Parking configuration and access: the Legacy parking structure is a single, efficient, one -level structure, which is very cost effective to build. It would benefit from an access point at the northwest corner, but this appears possible based on conceptual site plan. Design recommendations Legacy Reconfigure retail space around the plaza: we think the retail areas could be increased, and that the east side of the plaza should be re- worked. The retail behind the community room is dysfunctional, and the live -work spaces would be better suited for retail. Move the police: While it is a great community amenity, the leading edge of the plaza building is too valuable to give to the police station. This space should house the bellweather retail tenant in the project. The building could also be somewhat larger without sacrificing the visibility into the plaza. FULLER -SEARS Tukwila Village Developer Evaluation ARCHIT RAVENHURST ECTS September 12, 2008 DEN ELOPN1E\T,INC. 5 Parking improvements: While we agree that the parking strip along TIB is appropriate, the access and dead -ends are impractical. Better traffic flow through that area can and should be designed. The internal surface parking can be much more efficient, and access into the garage can be improved. Smaller plaza: Even the smaller Legacy plaza may be too big, so some programmatic analysis should be done to evaluate its utility. Design Details Tarragon View blockage into site: the large retail building at the TIB street edge, while maximizing the quality retail space in the project, blocks the view into the large central plaza from the street, so the primary public amenity of the project cannot be seen from the perimeter. This also blocks the line of sight to the other retailers at the back of the project, which are behind the large building and at the far side of the large plaza. This space will be hard to lease. Corner retail issues, drug store issues: We assume that the 15,000sf tenant is a drug store, presumably Bartells relocated from across the street. If this is the case, it adds a healthy anchor tenant to the project, but creates a number of urban design problems. First, the drug store will not merchandise the street side of the project, and it will either become a large blank wall, or mandated storefronts with nothing in the windows. Second, the plaza side of the building, which will also have little glass, will be dominated by the drive through and the loading docks. This does not activate the plaza the way one would like. There is also the question whether it is in the city's best interest to create a problem in the existing shopping center by moving their anchor tenant into the new project. Outdoor plaza: the outdoor plaza has the potential to be a great amenity for the community, but at 20,000sf it is not a plaza, it is an urban park. As it is across the parking Tots from any of the retailers or the library, it is something of an island, and will need specific amenities to attract people to go there. The coffee shop as shown will not be sufficient, and in fact, the coffee retailer will likely want a more prominent location with more pedestrian traffic. A space on the library plaza, which is the best urban space in the project, would be much better for them. Parking access and configuration. The Tarragon parking garages are all above grade, multi -story structures. As such, they are somewhat visible and will FULLER-SEARS Tukwila Village Developer Evaluation f ARCHITECTS September 12, 2008 DEV 1 6 need to be architecturally treated, and they will need to have internal ramps, which make them far Tess efficient. The access from the parking (especially the second levels) to the library and to the residential lobbies is unclear and probably convoluted. Retail parking shortfall: The surface parking around the plaza is not sufficient to serve the retail spaces or the library, so some of the parking structures will be used by retail and public customers. Pedestrian access, garage controls and revenue collection becomes more difficult in a shared parking configuration like this. The retail parking for the shops south of 144 and facing on TIB is non- existent, and as a result, the space will be very hard to lease. Residential lobbies, sky bridges: the concept plan shows one residential lobby and elevator core for the two buildings on the north half of the site. This will require a multi -level sky- bridge to connect the buildings, which is totally impractical. There will need to be a lobby and elevator core near the library, which makes the parking distribution between the lots and the buildings more difficult. Pea patch roofs: the Tarragon program has an admirable amount of pea patches, landscape and other activity on the roofs of the retail buildings, but in reality, most of this will be too expensive to provide and maintain. 144 parking: The idea of adding angled parking in the 144 right -of -way is an excellent idea for the project, but we don't know whether it conflicts with the city's plans to upgrade that street. 3 story and 4 story residential floors: The lower building heights allow the project to fit within the existing neighborhood more comfortably, but they create very Iong.internal hallways and are an inefficient (ie: more expensive) footprint. Plaza grade: the site drops nine feet from the TIB intersection to the coffee shop in the plaza. This puts the plaza in a hole behind a big building, which is not conveyed in the attractive renderings. In the next design stage, the design team will have to reconcile the site plan to the existing topography. Library on street: the library is very prominent on TIB, and is built to the edge of the street. This is very attractive in terms of highlighting the library and making it visible to the thousands of cars driving by, but it is also a noisy location, and FULLER -SEARS Tukwila Village Developer Evaluation RAVENHURST ARCHITECTS September 12, 2008 DEVELOPMENT. INC. 7 some buffer to the street might be better. Similarly, the plaza between the library and the restaurant is a great space, but it will also suffer from street noise, and so is not the best place for outdoor dining. The site plan could easily be adjusted to resolve these issues while retaining the overall character of the project. Design Recommendations Tarragon Less retail: We believe there is too much retail space in the Tarragon plan, both from a market perspective, and in terms of quality retail space. Consolidate and shrink the plazas: we fear the central plaza will be inactive compared to the library plaza, so the open space could be consolidated closer to the retail stores. Rethink parking configurations: it appears the parking program and organization has not been given great thought. This program should be re- organized and balanced for the various uses it needs to serve. Parking is the economic bane of any urban project, so anything that can be done to reduce cost must be considered. Move townhouses, change to retail: the space shown for townhouses on the south side of 144 is actually fairly good retail space. Four townhouses in the middle of the project seem out of place, and serve only to block the view of the garage. 4. Economic Deal Structure Overview As was stated in the introduction, we were provided very little economic information from Tarragon, and incomplete information from the Legacy proposal. We were told by Derek Speck that Tarragon has suggested the land price be based on an appraisal, while Legacy proposed a price based on a specific set of project economics. Given these circumstances, we have taken two approaches to the economic evaluation of the proposals. The first is simply the consideration of some of the deal points and project elements that have an impact on the economic structure. The list is not comprehensive, but it introduces some elements that can form the basis for further discussion between the City and the potential developers. FULLER Tukwila Village Developer Evaluation ARCHITECTS September 12, 2008 R a E 1 HURST 8 The second approach is to develop full development pro formas (project budgets and income statements) for each phase of each proposal, based on the physical design and facilities program they each propose, applying some of the proposal data and using industry standards and our own experience to flesh out the details. Then by evaluating each program based on its hypothetical performance, and by comparing the programs to one another, we can get some sense of the relative strengths of the two proposals. Legacy program: notable economic features. Higher land price: the Legacy land price of $5,385,000 is $3 million higher than the figure Derek uses for Tarragon, but that might be mitigated by the Tarragon offer of appraised value. On the face of it, it appears that Legacy is pushing harder to make the best offer they can. Separate senior and townhouse ownership: a key difference between the Legacy and Tarragon offers is that Legacy proposes to sub divide the site and carve off two pads for LIHI, one for a senior housing building, and one for a for sale townhouse project. The senior housing land is valued at $16,000 per unit, and the townhouse land is at $29,000 per unit, and the total LIHI land contribution is $1,520,000. Without checking market comps, these figures appear reasonable. But the importance is not in the land value; it is that by bringing in an affordable housing specialist that can build distinctly different product, Legacy is reducing the overall risk in the project while increasing the likelihood that the elements will get built. By including these elements, the number of market -rate rental units goes down, thus improving the odds that the market -rate housing can be absorbed. Neither Legacy nor Tarragon, to our knowledge, have specialized expertise for either senior or affordable for -sale housing, but LIHI does. We consider this a very smart move. More in Phase 1: Because of the sale to LIHI, which includes about half of the land south of 144 the Legacy program delivers more in Phase 1 and leaves less to Phase 2. We think the city should focus their evaluation on Phase 1, because if Phase 1 is not successful, there will be no Phase 2. Parking swap complications: The Legacy plan calls for a and swap with the apartments to the northeast, moving some of their parking on to the library site. This may be a very good idea- it appears to create a more efficient garage floor plate, but it also introduces a complication and another agreement that is FULLER -SEARS Tukwila Village Developer Evaluation RAVENHURST ARCHITECTS September 12, 2008 DEVELOPMENT, INC. 9 necessary for the project to move forward. Legacy should be quizzed on their level of confidence that this land trade can be achieved. Community room economics: The Legacy program includes a 3400 sf community room as a fully enclosed interior space. This is a significant public amenity, and a very expensive feature that drags down the financial performance of the project. Its fully allocated cost to the project probably approaches $1 million, so the city should carefully consider whether they have a real use for the space, or whether it should be deleted in exchange for elements of higher priority. If there is a realistic program of uses for the room, it could contribute significantly to the civic character of the project. Library expansion: The Legacy program also includes 2400 sf of library expansion space. There is no indication in the RFP that this was requested, and it appears to be left-over space in the conceptual plan. As with the community room, the City should decide whether this space is needed, and if not, eliminate it to improve the project economics. Tarragon program: notable economic features. Lower land price: the materials provided by Derek Speck suggest a land price of approximately $2,352,000, which we used in the pro forma analysis. This figure is $3 million less than the figures suggested by Legacy. If in fact Tarragon has proposed an appraisal approach to the land valuation, this differential may be meaningless, but based on the numbers provided to us, it is the most significant difference between the programs. We have not done an appraisal of the property. Lots of retail: The Tarragon program has a very significant retail program with over 50,000sf of shops and restaurants. This helps their pro forma, as the retail performs better than the residential, but we have major concerns that the market may not be able to support a retail project of this size. Even if a retail anchor of 15,000sf can be secured, (see below) this still leaves nearly 40,000sf to be leased. Using an average tenant size of 2000sf, which is large for a typical small shop retail tenant, Tarragon will have to find another twenty tenants for the project. We believe this site is a neighborhood retail location, similar to the retail located across the street, and the market is not deep enough to fill all the space that is proposed. If Tarragon proposes the project have a different and distinct retail character, they should be challenged to describe specifically what that is. FULLER SEARS Tukwila Village Developer Evaluation ARCHITECTS September 12, 2008 RAVEiVHURST DEVELOPMENT, LrC. 10 Having said that, Tarragon confounded the "experts" when theft Kent Station project was so successful, so their retail expertise should not be discounted. But it would be wishful thinking to expect a Kent Station program on this site, because of the proximity of the Southcenter retail district and the very different area demographics. The retail program is what makes the Tarragon proposal attractive, but there is a very high risk that this vision cannot be achieved. Low apartment rents: Based on the information provided to us, it appears that Tarragon is projecting apartment rents of $1.35/sf, and an average unit size of 865sf. This may be exactly what the local market needs and can support, but rents of $1.35 will not pay for a project with structured parking, extensive public space and subsidized community tenants. Given current construction costs, we would expect something closer to $1.85, which is also higher than the $1.65 that Legacy uses in their analysis. So both programs require some subsidy, but the gap to be filled if the rents are actually $1.35 is greater than anyone, public or private, can afford. Larger apartment units: The average unit size is 155 sf larger than Legacy's units, but it is unclear whether Tarragon is quoting gross or net area, so this may be a non issue. If the market for housing is primarily for immigrant families, larger unit sizes may be appropriate. If the target market is fully- employed young urban professionals, a smaller average unit size would be more appropriate. Anchor tenant: the Tarragon plan shows an anchor tenant of 15,000sf on the corner of TIB and 144 We agree that a project with over 50,000sf needs an anchor, and speculate that it is intended to be Bartells. If not, the list is very short for tenants of that size. Affordable housing program: The Tarragon proposal describes an affordable housing program, but does not discuss any special financing mechanisms for that component. We assume that both developers would use the tax abatement program that was offered, but Tarragon may need additional funding programs if their affordable housing program is significant. (Legacy solved this by selling land to LIHI.) When evaluating both development programs, the city should compare market rent levels to the income restricted rent levels. It is quite possible that market rents in this neighborhood are "affordable and if that is the case; you should determine whether the goal is to provide affordable housing that is below market rents. FULLER-SEARS Tukwila Village Developer Evaluation ARCHITECTS September 12, 2008 RAVENHURST DEV ELOPME \T, INC, 11 Phase 1 Parking: It appears there are approximately 285 parking stalls in Tarragon's Phase 1, but the parking needs are more likely to be nearly 400 stalls (residential at 1.3 stalls per unit, and retail and the library at over 3 stalls per 1000 sf). Adding these stalls would increase costs by $1 to $2 million, without generating significant revenues. Police station size: Just as the Legacy program has a larger community room and library, the Tarragon program adds 230 sf over the police space requirements. The number is minor, but it does add to the features that distinguish the two plans. 5. Pro forma Financial Analysis In order to evaluate the financial performance of the two proposals, we prepared complete business projections for each of the two programs. The models consist of projected capital budgets for each project, and pro forma income statements, with revenues and expenses. An assumptions page is used to collect inputs, and a summary page is used to describe the results. To prepare the models, we used the facilities program described by each developer, so the mix of uses, areas, parking counts and public elements come from each proposal. Neither package provided complete cost information, so we used our own experience in assembling capital budgets and development costs for each one. We used the apartment rent figures provided by each developer, but used equal retail rents and library/police station rents. The purpose was to arrive at an order of magnitude financial assessment for each plan, and to provide a platform for making adjustments to better compare the two plans. In the end, we developed three separate models. We modeled the Legacy program as it was presented in the response to the RFP. We modeled the Tarragon program as it was presented in the RFP, but we had to fill in some of the blanks. And because the Tarragon plan is based on significantly different apartment assumptions, we ran a second Tarragon model using the Legacy rent rates, unit sizes and land price. Base Model: Both Programs as Proposed. As the two projects were presented in the responses to the RFP, the costs and returns are as follows: FULLER `SEARS Tukwila Village Developer Evaluation ARCHITECTS September 12, 2008 RAVENHURST DEVELOP IE \T. INC 12 Total Project Cost Phase 1 Phase 2 Consolidated Legacy: 46,476,000 16,655,000 63,132,000 Tarragon: 42,488,000 39,667,000 82,155,000 The total project costs for Tarragon are higher than for Legacy, because the Legacy costs do not include the LIHI housing components, and Tarragon builds more retail space. It should be noted that we subtracted the LIHI land payments of $1,520,000 from the Legacy land cost figures, as these funds will basically flow through Legacy to the City and as such are not a Legacy project cost. Legacy builds more in Phase 1, with a small Phase 2. Tarragon builds more than half of its housing in Phase 2. Either project is a very significant undertaking, and of a project size for which only the strongest developers can find financing today. Return on Cost: Phase 1 Phase 2 Consolidated 'Legacy: 6.7% 6.8% 6.7% Tarragon: 6 .9% 5.7% 6 .3% Return on cost is the simple measure that developers use to determine whether the project makes money. The return on cost needs to be higher than the interest rate on debt (6.5% in this case), and should be higher than the returns a purchaser would expect if they were to buy the property on completion (probably 7.0% to 7.5% today). Neither project performs well enough to meet these thresholds, although they are close. You should expect the final development pro forma to show a residential return on cost of at least 6.5 and at least an 8.5% return on cost for the retail, for a combined 7.5% return. For either project to go forward, something good needs to happen. It is possible that our cost estimates are high, and reality is a little less pessimistic. We have assumed fairly low rents for both the apartments and the retail. Other new projects around Seattle have needed retail rents of over $30 /sf /yr and residential rents of at least $1.85 /sf /mo in order to be financially successful, and applying these rent levels to the FULLER-SEARS Tukwila Village Developer Evaluation RAVENHURST ARCHITECTS September 12, 2008 DEYELOPNI ENT, INC 13 current models produces returns for both Legacy and Tarragon that would be satisfactory in the current market. But one cannot just assume higher rents, and we expect the developers will be challenged to achieve the more conservative numbers we use in the models, which is $25 retail rents and $1.65 for the residential rents. The Tarragon base model uses residential rents of only $1.35, and it is highly unlikely the project makes any economic sense at those numbers. Second Model: Equivalent Rents and Land Costs. For a more equitable comparison„ we applied the Legacy rent figure ($1.65 /sf /mo for the apartments) and the Legacy unit size (750 sf net versus 865 sf net) to the Tarragon development program. We also adjusted the land payment to the city so that it was the same for both developers. This produced the "Tarragon Apples to Apples" model, which compares to the Legacy program as follows: Total Project Cost Phase 1 Phase 2 Consolidated 46,476 000 16,655,000 63,13, Legacy: 63,132,000 Tarragon Apples: 41,377,000: 37,148,000 78,525,000 The cost of the Tarragon project is reduced by $4 million even though the land price goes up, because the residential unit sizes are reduced. Return on Cost: Phase 1 Phase 2 Consolidated Legacy: 6.7% 6.8% 6.7% :Tarragon A pies: 7 5 °Y 6.6 7.1% The return on cost of the Tarragon project goes up because the residential rents go up, which more than offsets the increase in land price. At the end of the day, the Tarragon Apples to Apples project outperforms the Legacy program because of the retail component. The Legacy site has more residential units (including LIHI), but little retail, while the Tarragon design has a large retail element. But we have some concern that the large retail program is viable, and it may never fully lease up. FULLER -SEARS Tukwila Village Developer Evaluation Hi ARCHITECTS September 12, 2008 RAVENHUR5T DEVELOPMENT, INC. 14 Third Model: Retail Vacancy Adjustments. If you were to assume a higher vacancy rate for the Tarragon plan because the market won't support that much retail, which we believe but cannot support without a retail market study, the two projects look very similar. Using a 10% retail vacancy rate for the Legacy plan and a 25% vacancy rate for the Tarragon program, the two models are virtually identical: Return on Cost: Phase 1 Phase 2 Consolidated Legacy: 6.7% 6.8% 6.7% ;Tarragon 25% Apples: 6.9% 6.4 6.7% We therefore believe the choice of the two developers cannot and should not be made based on a review of the financial models. The assumptions used in the two proposals were very different, and the number of variables can make your head spin. The four models (adding the "Tarragon 25% Apples" to the list) can be found in an appendix at the end of this report. 6. Developer Experience Experience is an excellent and appropriate criterion for the developer selection, but in this case, both developers are capable, financially stable and innovative, and either one would be a good choice. Legacy Partners: Legacy's current projects tend to be more urban and of larger scale than the Tukwila project. Legacy is primarily a residential developer, but virtually all of their projects include some amount of commercial, and they are fully qualified to develop an urban mixed -use project. Their Redmond project currently under construction is a good example of an innovative, multi- block, mixed -use property. While based in California, their Mercer Island office has ample horsepower to execute a project of this scale. Legacy's greatest strengths are there extensive residential development background, and a very pragmatic, responsible approach to development. Tarragon: Tarragon is a more diverse and Tess- focused developer. Their portfolio includes industrial development and suburban shopping centers as well as some very innovative urban mixed -use projects. They do not have the residential expertise that FULLER SEARS Tukwila Village Developer Evaluation ARCHITECTS September 12, 2008 RAVENHL�RST p DEVELOPMENT. INC. 15 Legacy has, but they have become strong retail developers. Their most comparable project is Saffron. They are opportunity- driven and undertake challenging projects that appeal to them. Their staffing levels and development capacity are similar to Legacy's. 7. Project Risks Retail risk: we believe this is a very difficult retail site. It primarily serves a neighborhood retail trade area with low market rents, and the tenant types are not significantly different than the uses already found in the neighborhood. There may be an untapped market opportunity, but if so, it may be an ethnic one, not a cosmopolitan, glossy retail concept. We fear that the city's ambition for the site is more visionary than the market can support, and pursuing that vision will delay a more realistic project that can ultimately be built. Housing risk: we believe the housing is the economic engine that will drive the successful development of this site. We suspect that current market rents will not support the construction types envisioned for the project by either of the two developers, and that careful and efficient design and well- executed affordable housing components will be necessary for the project's success. Structured parking is the element that often kills a project's economics, so a very rational parking program will be needed. Legacy's biggest risks: Legacy's program, as proposed, will be challenged by residential market rents, by the economic drag of the community space, by the added risk of the parking lot trade, and by the questionable need for library space expansion. Some of these can be mitigated by further design and program refinements. Tarragon's biggest risks: Tarragon's program, as proposed, will be severely challenged by the visionary but very aggressive retail plan, by the economic drag of inefficient parking garage design, by residential inefficiencies and excessive amenities, and by the plan's reliance on parking along 144 Some of these risks will be eliminated in further design, but we expect the result will be a residential and parking plan that looks more like the Legacy program. FULLER -SEARS Tukwila Village Developer Evaluation f ACECTS September 12, 2008 DAB oNHGRST RHET 16 8. Our Ideal Project Mix: Our experience tells us the ideal project composition will be a program that includes 20 to 25,000sf of retail in addition to the library space, with multiple housing types, developed in three or four phases, serving moderate to middle incomes. The project will depend on multiple funding sources as is typical for affordable housing. Structured parking will be simplified into an efficient plan, with more surface parking, efficiently laid out, than either plan currently shows. The community room will shrink, as will the plazas, and they will be consolidated to the hot spot on the site where the most activity is located. 9. Questions for Each Developer Before making a selection, we would recommend you ask the developers a few more questions. For Legacy you should ask the following: What is your justification for the $1.65 apartment rents? Can you elaborate on the size and mix of units, and on the target market for these apartments? Would you be willing to increase the retail component of the project? What is your plan for the retail mix and uses? For Tarragon, you should ask the following: How did you arrive at your $1.35 rents for the residential units? Can you make this project pencil at those rent numbers? What is the Bartells level of interest? If Bartells is not your potential anchor tenant, do you have other realistic candidates for the space? What is the retail merchandising concept? What kinds of tenants do you have in mind for the space? Do you expect to pay more for the land than is in your proposal? What are your realistic land price expectations? Attachments: FS Design Review, Pro formas FULLER Tukwila Village Developer Evaluation ARCHITECTS September 12, 2008 RAVENHtsRST DEVELOPME \T Li C. 17 RAVENHURST DEVELOPMENT, INC. DRAFT 9/12/2008 Tukwila Village Legacy CONSOLIDATED PROFORMA PROGRAM SUMMARY Phase 1 Phase 2 Consolidated Library sf 10,000 2,400 12,400 Shops sf 12,000 6,500 18,500 Total Retail GLA sf 25,400 8,900 34,300 Apartment Units units 200 68 268 Parking stalls 380 80 460 COST SUMMARY: Land $2,820,000 3,866,000 Offsites $150,000 $0 150,000 Construction: $34,935,458 $12,093,146 47,028,603 Design Fees $2,311,127 $850,589 3,161,716 Development Costs $2,101,773 $814,657 2,916,430 Leasing and Marketing $300,000 $125,000 425,000 Tenant Allowances $220,000 $65,000 285,000 Interest and Financing $2,691,143 $1,068,287 3,759,430 Contingencies $946,645 $592,762 1,539,407 TOTAL COSTS: $46,476,146 $16,655,441 63,131,587 OPERATIONS: Net Operating Income, Commercial: $441,318 $214,016 $655,334 Net Operating Income, Apartments: $2,681,291 $919,307 $3,600,598 Net Operating Income, Parking: $0 $0 $0 Total Net Operating Income: $3,122,610 $1,133,323 $4,255,932 Debt Service 6.50% 30 yr. Amort ($2,562,016) ($619,907) ($3,181,923) NET CASH FLOW: $560,594 $513,416 $1,074,009 RETURN ON COST: 6.7% 6.8% 6.7% PROJECTED PROJECT VALUE (year 3) $44,608,708 $16,190,327 $60,799,034 POTENTIAL PROFIT (year 3) ($1,867,438) ($465,115) ($2,332,553) SALE CAP RATE 7.00% 7.00% 7.00% CONSTRUCTION EQUITY: $16,266,651 $5,829,405 $22,096,056 CONSTRUCTION LOAN AMOUNT $30,209,495 $10,826,037 $41,035,532 CONSTRUCTION LOAN TO COST 65.0% 65.0% 65.0% CONSTRUCTION LOAN TO VALUE 67.7% 66.9% 67.5% PERMANENT EQUITY: $13,019,615 $4,512,696 $17,532,311 PERMANENT LOAN AMOUNT $33 ,456,531 $12,142,745 $45,599,276 PERMANENT LOAN TO COST 72.0% 72.9% 72.2% PERMANENT LOAN TO VALUE 75.0% 75.0% 75.0% PERMANENT DEBT COVERAGE RATIO 1.22 1.83 1.34 RETURN ON PERMANENT EQUITY (YR 3): 4.3% 11.4% 6.1 Drawing Reference: none Contractor Estimate: none legacy pf consol Tess LIHI 080912 .xlsx summary 9/12/2008 1:21 PM Pagel RAVENHURST DEVELOPMENT, INC. DRAFT 9/12/2008 Tukwila Village Tarragon CONSOLIDATED PROFORMA PROGRAM SUMMARY Phase 1 Phase 2 Consolidated Library sf 10,000 0 10,000 Shops sf 43,985 12,715 56,700 Total Retail GLA sf 53,985 12,715 66,700 Apartment Units units 144 158 302 Parking stalls 285 284 569 COST SUMMARY: Land $1,410,000 $940,000 $2,350,000 Offsites $0 $0 $0 Construction: $32,576,423 $31,743,093 $64,319,517 Design Fees $2,169,585 $2,004,586 $4,174,171 Development Costs $1,983,821 $1,701,155 $3,684,976 Leasing and Marketing $498,040 $264,720 $762,760 Tenant Allowances $539,850 $127,150 $667,000 Interest and Financing $2,551,041 $2,431,084 $4,982,124 Contingencies $759,565 $454,881 $1,214,445 TOTAL COSTS: $42,488,325 $39,666,668 $82,154,993 OPERATIONS: Net Operating Income, Commercial: $1,131,225 $336,699 $1,467,924 Net Operating Income, Apartments: $1,804,241 $1,920,521 $3,724,762 Net Operating Income, Parking: $0 $0 $0 Total Net Operating income: $2,935,466 $2,257,220 $5,192,686 Debt Service 6.50% 30 yr. Amort ($2,408,469) ($1,234,658) ($3,643,127) NET CASH FLOW: $526,997 $1,022,562 $1,549,559 RETURN ON COST: 6.9% 5.7% 6.3% PROJECTED PROJECT VALUE (year 3) $41,935,228 $32,246,001 $74,181,229 POTENTIAL PROFIT (year 3) ($553,098) ($7,420,667) ($7,973,764) SALE CAP RATE 7.00% 7.00% 7.00% CONSTRUCTION EQUITY: $14,870,914 $13,883,334 $28,754,248 CONSTRUCTION LOAN AMOUNT $27,617,412 $25,783,334 $53,400,745 CONSTRUCTION LOAN TO COST 65.0% 65.0% 65.0% CONSTRUCTION LOAN TO VALUE 65.9% 80.0% 72.0% PERMANENT EQUITY: $11,036,905 $15,482,167 $26,519,072 PERMANENT LOAN AMOUNT $31,451,421 $24,184,501 $55,635,921 PERMANENT LOAN TO COST 74.0% 61.0% 67.7% PERMANENT LOAN TO VALUE 75.0% 75.0% 75.0% PERMANENT DEBT COVERAGE RATIO 1.22 1.83 1.43 RETURN ON PERMANENT EQUITY (YR 3): 4.8% 6.6% 5.8% Drawing Reference: none Contractor Estimate: none tarragon of consol 080903 .xlsx summary 9/12/2008 1:22 PM Pagel RAVENHURST DEVELOPMENT, INC. DRAFT 9/12/2008 Tukwila Village Tarragon Apples to Apples CONSOLIDATED PROFORMA PROGRAM SUMMARY Phase 1 Phase 2 Consolidated Library sf 10,000 0 10,000 Shops sf 43,985 12,715 56,700 Total Retail GLA sf 53,985 12,715 66,700 Apartment Units units 144 158 302 Parking stalls 285 284 569 COST SUMMARY: Land $3,477,000 $1,908,000 $5,385,000 Offsites $0 $0 $0 Construction: $29,893,531 $28,799,363 $58,692,894 Design Fees $2,008,612 $1,827,962 $3,836,574 Development Costs $1,849,677 $1,553,968 $3,403,645 Leasing and Marketing $498,040 $264,720 $762,760 Tenant Allowances $539,850 $127,150 $667,000 Interest and Financing $2,365,659 $2,227,679 $4,593,339 Contingencies $744,809 $438,690 $1,183,499 TOTAL COSTS: $41,377,177 $37,147,533 $78,524,710 OPERATIONS: Net Operating Income, Commercial: $1,131,077 $336,699 $1,467,776 Net Operating Income, Apartments: $1,979,700 52,129,890 $4,109,590 Net Operating Income, Parking: $0 $0 $0 Total Net Operating Income: $3,110,777 $2,466,589 $5,577,366 Debt Service 6.50% 30 yr. Amort ($2,382,154) ($1,349,179) ($3,731,333) NET CASH FLOW: $728,623 $1,117,410 $1,846,033 RETURN ON COST: 7.5% 6.6% 7.1% PROJECTED PROJECT VALUE (year 3) $44,439,672 $35,236,991 $79,676,663 POTENTIAL PROFIT (year 3) $3,062,495 ($1,910,542) $1,151,953 SALE CAP RATE 7.00% 7.00% 7.00% CONSTRUCTION EQUITY: $14,482,012 $13,001,637 $27,483,649 CONSTRUCTION LOAN AMOUNT $26,895,165 $24,145,896 $51,041,062 CONSTRUCTION LOAN TO COST 65.0% 65.0% 65.0% CONSTRUCTION LOAN TO VALUE 60.5% 68.5% 64.1% PERMANENT EQUITY: $10,269,407 $12,481,639 $22,751,046 PERMANENT LOAN AMOUNT $31,107,771 $24,665,894 $55,773,664 PERMANENT LOAN TO COST 75.2% 66.4% 71.0% PERMANENT LOAN TO VALUE 70.0% 70.0% 70.0% PERMANENT DEBT COVERAGE RATIO 1.31 1.83 1.49 RETURN ON PERMANENT EQUITY (YR 3): 7.1 9,0% 8.1 Drawing Reference: none Contractor Estimate: none tarragon pf consol apples 080903 .xlsx summary 9/12/2008 1:22 PM Pagel RAVENHURST DEVELOPMENT, INC. DRAFT 9/12/2008 Tukwila Village Tarragon Apples 25% CONSOLIDATED PROFORMA PROGRAM SUMMARY Phase 1 Phase 2 Consolidated Library sf 10,000 0 10,000 Shops sf 43,985 12,715 56,700 Total Retail GLA sf 53,985 12,715 66,700 Apartment Units units 144 158 302 Parking stalls 285 284 569 COST SUMMARY: Land $3,477,000 $1,908,000 $5,385,000 Offsites $0 $0 $0 Construction: $29,893,531 $28,799,363 $58,692,894 Design Fees $2,008,612 $1,827,962 $3,836,574 Development Costs $1,849,677 $1,553,968 $3,403,645 Leasing and Marketing $498,040 $264,720 $762,760 Tenant Allowances $539,850 $127,150 $667,000 Interest and Financing $2,365,659 $2,227,679 $4,593,339 Contingencies $744,809 $438,690 $1,183,499 TOTAL COSTS: $41,377,177 $37,147,533 $78,524,710 OPERATIONS: Net Operating Income, Commercial: $884,962 $258,715 $1,143,677 Net Operating Income, Apartments: $1,979,700 $2,129,890 $4,109,590 Net Operating Income, Parking: $0 $0 $0 Total Net Operating Income: $2,864,662 $2,388,605 $5,253,268 Debt Service 6.50% 30 yr. Amort ($2,350,377) ($1,306,524) ($3,656,901) NET CASH FLOW: $514,285 $1,082,081 $1,596,367 RETURN ON COST: 6.9% 6.4% 6.7% PROJECTED PROJECT VALUE (year 3) $40,923,746 $34,122,935 $75,046,680 POTENTIAL PROFIT (year 3) ($453,432) ($3,024,598) ($3,478,030) SALE CAP RATE 7.00% 7.00% 7.00% CONSTRUCTION EQUITY: $14,482,012 $13,001,637 $27,483,649 CONSTRUCTION LOAN AMOUNT $26,895,165 $24,145,896 $51,041,062 CONSTRUCTION LOAN TO COST 65.0% 65.0% 65.0% CONSTRUCTION LOAN TO VALUE 65.7% 70.8% 68.0% PERMANENT EQUITY: $10,684,368 $11,555,332 $22,239,700 PERMANENT LOAN AMOUNT $30,692,809 $25,592,201 $56,285,010 PERMANENT LOAN TO COST 74.2% 68.9% 71.7% PERMANENT LOAN TO VALUE 75.0% 75.0% 75.0% PERMANENT DEBT COVERAGE RATIO 1.22 1.83 1.44 RETURN ON PERMANENT EQUITY (YR 3): 4.8% 9.4% 7.2% Drawing Reference: none Contractor Estimate: none tarragon pf consol apples 25% 080903 .xlsx summary 9/12/2008 1:22 PM Pagel