LLAMA is a present value cash flow model. LLAMA calculates the cash flow from loans or groups of loans over the life of the loans. The cash flow is discounted to provide a present value of the cash flows. The resulting dollar value is often referred to as the "value of the portfolio".

The LLAMA model does one thing, it creates a value for a loan or group of loans using a previously created set of assumptions.

1. You get full control of the valuation process.
2. No more expensive updates
3. Unlimited loan counts, segments, or portfolios.
4. Direct control using Microsoft's Visual Basic
5. Direct control of all input screens
6. Custom designed output reports.
7. Unlimited flexibility to redesign the cash flow process.
8. You get the source code!
9. Full life of the loan vectoring and assumption tables.
10. User defined investor remittance patterns.

There are two basic forms of assumptions:

1. An array of values for 360 month or more,

2. A single value.

These assumption vectors can be created in any program you desire. You only need put one value per line for 360 months. This gives you complete control over the assumptions for valuation. Delinquencies, fees and costs can all vary for the full 360 mont hs.

The model takes the single input file and merges the data using an assumption table that you create. The assumption table key can be anything you want. This key must be on each record. If the key is FNMA then there must be a unique record in the assump tion table for "FNMA".

The standard data file is a flat ASCII file that contains the information described in VA01D. It can be a summary file or one loan per record. The model calculates the remaining ratios and amortization schedules based on a single loan and then multiple s these values times the number of loans in the segment. As long as the loans in the segments have similar characteristics the value will be correct.

You can easily reprogram the LLAMA model in any computer language. There is a single input, calculation and output. This is the reason why all vectors are kept as simple ASCII files. The model is easily adapted to your database environment.

Inputs - The model uses a standard set of inputs that can be supplied in any format database, ASCII or even spreadsheet. It will run in any operating system and with and stratification system.

1. Data table
2. Assumption set - This file directs the model what to do and which vector files to use. These can be maintained and used from month to month.
3. Vector files - These files define a rate or factor for the length of the valuation. They can be created using any word processor or spread sheet

All inputs, assumptions, outputs, tables and other results are made available directly to the user. A complete valuation application can easily be designed from any database or computer application.

No attempt is made to provide an OAS value. This is only a cash flow model. It is possible to adapt the model to quickly run multiple interest rate scenarios.

LLAMA reads fixed length files or tables of information and calculates a present value for the loans or segments in those files. It is calls "engines" because that is the only piece of the system that is sold to the user. All inputs and outputs are com pletely defined in the documentation. Each user can build their own interface to the engine. This assumes that each user already has a database of information. Rather than duplicate the database functions, LLAMA simply provides that calculation engine to value those loans or segments.

The engine is simply a set of calculations that are easily adapted to any processing platform. No specific computer formats or database systems are required.

Client Data - The client data is usually available in a mainframe format.

Data input files - The basic data input file is a fixed length ASCII file which contains all information need to value the loan.

Formats - ASCII, DBF, TEXT, fixed - Although I usually refer to the input file as fixed ASCII, the file can have any format that is PC compatible. If a different DOS is used, such as UNIX, then the engine will provide those formats.

Data conversion module - LLAMA creates the data record using client information.

Segment data - LLAMA will value single loans or groups of loans. The engine is built to process sequential files of records. There can be an unlimited number of loans in the file. Also, additional files of loans can be included in the valuation.
Group results - LLAMA will accumulate groups of cash flows during the valuation. You may have specific loan groupings that you wish to value. The engine accumulates the cash flows based on a data field that you put in the input record.