This is the first of a series of papers that the author is writing on models that can improve profitability in the securities lending business. We consider how these methodologies can be applied to both buy side and sell side institutions. We look at different models to manage spreads on daily securities loans and aid the price discovery process, improve the efficiency of the locate mechanism and optimize the allocation of inventory, develop strategies for placing bids on exclusive auctions, price long term loans as a contract with optionality embedded in it and also look at ways to benchmark which securities can be considered to be more in demand or highly shorted and use this approach to estimate which securities are potentially going to become “hot” or “special”, that is securities on which the loan rates can go up drastically and supply can get constrained. In this paper, the objective is either to design an appropriate securities lending auction mechanism or to come up with a strategy for placing bids, depending on which side of the fence a participant sits. There are two pieces to this puzzle. One is the valuation of the portfolio being auctioned subject to the information set available to the bidder or the auction designer. This information set would include among other things, the demand for the securities, any additional demand from the locates received, the loan rates applicable to those securities, the duration of the loans, the frequency of loan turnover and the internal inventory pool available to the bidder. These variables are modeled as geometric brownian motions with uncertainty introduced via suitable lognormal distributions and a symmetric normal distribution. We derive heuristics to arrive at a set of valuations, with a pecking order that can help decide the aggressiveness of the valuation.The other piece would be to come up with the best strategy from an auction perspective once a valuation has been obtained. We start with the benchmark scenario where the buyers, placing bids are assumed to have perfect and complete information regarding their valuation of the portfolio that is being auctioned, that is private only to them. We consider the uniform distribution as the simplest scenario and extend that to a more realistic setting that considers the valuations to be log normally distributed. We further extend this by introducing uncertainty into the estimation of bidder valuations and their bidding strategy. The possibility of number of bidders being unknown, the valuations from various bidders being correlated or the interdependent valuation framework and, a reserve price set by the auction seller are more complex extensions. It is easily seen based on existing results that the strategies of the bidders constitute a Nash equilibrium, under suitable conditions.Lastly, we run simulations to establish numerical examples for the set of valuations and for various bidding strategies corresponding to the different auction settings. It is tempting to call this one of the more (most) challenging problems in finance, and even though this is debatable and perhaps even labelled as due to ignorance on the author's part, what stands true is that this is certainly one of the least explored yet profit laden areas of modern investment management. The models developed here could be potentially useful for inventory estimation and for wholesale procurement of financial instruments and also non-financial commodities.