Executive Summary: The algorithmic dealer simulation is a simulation that places us in the shoes of a dealer trading in a pure dealer market. The objective is to create strategies that encourage and strengthen order flow. To do this we are asked to create trading strategies that cover a wide array of unique trading circumstances. Our goal is to provide strategies that maximize our average utility through all 60 problems. To design our strategies we took a broad approach and assigned each of our 5 team members to come up with 6 unique strategies. Some of our strategies were derived from the initial ten strategies and we built on them or tweaked them slightly to cover a wider range of circumstances. The remainder of our strategies were derived from past experiences of TraderEx simulations. In these instances, we took prior knowledge of what we did in those simulations and brought over some successful strategies and converted them into a format suitable for the algo dealer simulation. The diversity amongst our strategies allowed us to run the simulation with an average utility of ~197 with a maximum utility of 689. As we ran our simulations a few trends emerged that our group noticed. The first trend we observed dealt with the easiest scenarios, which we defined as being the scenarios that had low volatility, adverse selection, and risk aversion. Amongst these strategies we found that the best strategies often gave us small spreads at a competitive price which drove up our order flow and we profited through high volume. The strategies that were optimal in these instances almost always had an ask price that matched competitors prices or even slightly lower which allowed us to take the most market orders and profit through high order traffic. When we began to run the simulations on problems that had a higher risk aversion parameter, we observed that low quantity, orders that filled only quoted depths worked best. We also observed that many of our optimal strategies were inventory management focused. In short, the best strategies for these types of problems were low quantity order based with cautious execution tactics. When we noticed problems that had higher daytime volatility, we observed these same types of strategies also worked. When the overnight volatility factor increased, we asked ourselves how that might affect us as dealers. We concluded that to combat overnight volatility we would want as little inventory as possible at the end of the day to avoid adverse price movements overnight. In hindsight, we should’ve produced strategies to solve this problem through the use of IF statements dealing specifically with inventory. For example: if a problem had an extremely high overnight volatility factor, we should have implemented a strategy that had an embedded IF statement instructing us to sell at the lowest market price with maybe 3 periods left IF our inventory was above 10. Say our inventory at 3:00 PM was 100, we should have a strategy that would immediately lower our ask price to below our competitors in an attempt to sell this excess inventory to avoid negative price trends that may occur overnight. We also found that strategies that had a large spread were able to overcome some deficiencies the strategy had. Also, the large spread gave the strategy the ability to overcome the obstacles the problems sets presented. For example our strategy 12 was optimal despite having high adverse selection. The wide spread allowed the strategy to overcome this deficiency in the strategy and the strategy was still able to be an optimal strategy for problems 51-60. The problem set 51-60 also had a high order processing cost, which obviously would mean the strategy needed to have a smaller amount of trades to be effective.