Ryan Peng ’14, Hanweck Associates

Monday, March 19, 2012

I arrived at Hanweck Associates a couple of minutes before 9 am. Dr. Hanweck introduced us to everyone in the office and explained a bit about what his firm does on a day to day basis, although each day is different and can be unpredictable! Shortly afterwards, the big news of Apple announcing a dividend payout to its shareholders reached the office, so most people were working on updating the databases of option pricing information. After everyone settled down into normal work, Wesley, the other Princetern, and I talked with Dr. Hanweck and started to analyze historical data on dividend payout dates and amounts. Using Excel, we compared the actual dates and amounts to predicted dates and amounts provided by two vendors in order to see if the vendors were doing a good job. This is important because Hanweck Associates buys this dividend prediction data from the vendors and uses it in their calculations and models. We worked on this for the rest of the day and decided to switch from Excel to C++ in order to parse the data more efficiently. Throughout the afternoon, we got to talk with the associates in the office and learn more about what projects they were working on – lots of cool stuff all going on at once! At the end of the day, we summarized what we found and talked with Dr. Hanweck and Andy before leaving the building at 6.

Tuesday, March 20, 2012

On the second day, we started off by finishing up details on the dividend payout comparisons. We found some interesting patterns and results in the data, so we noted those and passed them on to next person who would be using them. After lunch, we started on our second project with the help of Jeff and Dr. Hanweck. The goal of this project was to build a computer program that approximates the Heston model option pricing equations, using the results of an earlier academic that derived first and second order approximation with stochastic calculus. We decided to implement this in Java, and we worked on this till the end of the day. It was definitely a tough project for us since the equations were pretty complex, and the paper was hard to follow since we had to grasp some hard mathematical ideas in order to understand the first and second order approximations. By the end of the day, we had a functioning program that complied but gave the wrong numerical results. We chatted with Dr. Hanweck about what his company does and about different option pricing methods before we left the workplace at 6:30.

Wednesday, March 21, 2012

Ryan, fellow Princetern Wesley, and Dr. Jerry Hanweck

Today, we continued to work on the Heston model approximation. We started off by going through the entire program to trace what happens at each step and to make sure that we put in all the approximation formula components correctly. With some help from Dr. Hanweck and a few others, we located some bugs in our code and fixed them. Now, our code was getting closer to the right numbers but was still off by about one percent. We decided to go for lunch and brainstorm ideas on how to perform additional testing. Since we did not know for sure whether this was caused by bugs in our code or whether this error is inherent in the approximation itself, we decided to look in academic journals for numerical data and results done by others. After lunch, we found some data and compared it to our results and saw that we matched their results fairly closely. We finished up the program by cleaning up the variables and adding some extra comments so that the next person who works with this code can better understand what we were doing. It was getting close to the end of the day, and I couldn’t believe how fast this Princeternship went by! We looked around the workplace and talked with others to see what they were currently working on. Afterwards, we finished the day by chatting with Dr. Hanweck about what we had done with the Heston model approximation project. Overall, I greatly enjoyed this Princeternship and learned a lot through hands-on experience. I was able to apply some of the material I had learned in ORF 335: Financial Mathematics, so the practical experience nicely complemented the theory I had learned in the classroom. I would like to thank Dr. Hanweck, Andy, Jeff, and everyone else at Hanweck Associates who shared their knowledge and stories with us. It was a wonderful experience, and I am looking forward to working at a financial firm again.

John McNamara ’14, Hanweck Associates

My Princternship with Hanweck Associates gave me an interesting look at the intersection between technology and finance. I arrived at the office in lower Manhattan with my fellow Princetern, Eric, at about 8:15 am. Richard, one of the firm’s quantitative analysts, greeted us and gave us an introduction to the firm’s business and explained the technological focus of the firm as well as his own background and work at Hanweck. Soon after, we met our host Dr. Hanweck ’87. He then introduced us to Mike who gave us our first project. The project involved spin-offs, which, as Mike explained, involve a larger company splitting off a part of its business into a smaller, stand-alone company. We were tasked with researching spin-offs in the last decade and finding out how many shares of the new company someone received per share of the larger company they owned. After working on this for a few hours, we took a quick break for lunch. After lunch Dr. Hanweck invited us into his office to talk about his career, financial derivatives, high-speed trading technology, and his perspective on the future of his business and the industry as a whole. It was the first time I heard all the economics terms I have learned in class like price discrimination, consumer surplus, and comparative advantage used in a real life situation so it was an intriguing link between my studies and the business world. Aside from that, it was incredible to hear the perspective of an alumnus and learn about his career path and how he believes a business can be successful. After this conversation, I returned to work on the spin-offs project until Mike provided us with another project. He gave us a quick lesson on using queries in Excel and set us up for a project to make a user-friendly format to look up historic equity prices and make graphs from database information. After experimenting with Excel for a while and getting a feel for the database, it was 6:00 pm and time to head home. I took a subway and a train and was home by 8:00, and I was ready to see what the next day had in store for me.

January 31, 2012 

I began my second day at Hanweck Associates by working some more on the spin-offs project from the previous day. Afterwards, I returned to Excel and Microsoft Query and was able to set up a system in Excel where a user could plug in a stock ticker and immediately get data on the stock’s equity id number, valid dates, and other information from the company’s databases. Once Eric and I were comfortable with this, Mike provided us with objectives for some larger projects. He brought us into a conference room and gave us a lesson on options and their properties. Then he defined three projects for us to work on. One was to make an interface whereby a user could specify a stock ticker and a start and end date and see the stock information, the day-to-day closing price of the stock, and a graph plotting the equity price against time. The next was an option-contract table. It would allow the user to specify a certain equity-id number (associated with a stock) and a valid date and see all of the options associated with that equity id and valid date. The final project was to make a heat map based on the volatility indices of different options based on their “deltas” and time to expiration. I worked on the first two projects and learned a lot about how to manage databases. I also discovered trends in equity prices and option contracts as I tested the interfaces that I had created. After lunch, Eric and I spoke with Richard and learned about his current projects and the challenges of collecting, analyzing, and storing the incredible amounts of data that the firm collects in real time. I then returned to the projects I was working on and at the end of the day I had working systems in place for each. The projects taught me about the technological challenges in storing and retrieving large amounts of data. Before we left for the day, Dr. Hanweck spoke to us about a project for the next day that would allow us to compare dividend forecasts from two different companies the firm uses against the actual data to determine how close the predictions are. The day gave me some great hands on experience while learning more about financial derivatives and the markets in general.

February 1, 2012 

My final day at Hanweck Associates was centered on the dividends project that Dr. Hanweck had described to us the previous day. I arrived at the office early and after reading the financial news of the day, Eric and I met with Andy, a quantitative analyst at the firm. He explained the specifics of the project and the challenges of “unclean data.” It is often easy to assume that the data that we receive is well organized and accurate, but he explained that that is actually seldom the case. We were charged with looking at thousands of dividend payments and their respective payment dates along with the predictions for those dividend payments from two companies that the firm purchases forecast information from. Had the data been nice and clean, it would have been straightforward to write a simple program to calculate the forecast errors for each dividend for both of the companies’ predictions and then determine which one was better overall. Unfortunately, inconsistency in the data made this very difficult. Instead, we had to put the data in a consistent format and account for the gaps so that the data could be better analyzed. Once this was done, it was possible to compare the forecasts for the different companies and see which ones fared better both overall and in certain areas. Our day concluded with a final recap discussion with Dr. Hanweck. We discussed our experience and what we learned as he shared more of his story on how he started his own company and where he hopes the company will be moving forward. It was a great conclusion to the Princeternship as we had the chance to reflect on our experience and learn more about the firm and its CEO as well as the industry as a whole after having the opportunity to work on some projects and listen to conversations among its employees in the time we were there.

I thank Dr. Hanweck for this opportunity as well as all those in the office who helped and guided me in the time I was there. This is an excellent Princeternship for students interested in technology and finance as there is a lot to be learned from a technical standpoint as well as from the experience of working in the financial capital of the world alongside a group of dedicated and talented individuals.

Eric (Tiansheng) Guo ’14, Hanweck Associates

Fellow Princetern John, Dr. Jerry Hanweck, and Eric

Monday morning at 7:30 am, I met the other Princetern in the lobby of Twenty Exchange Place, where I stayed for all three nights of the Princeternship.  We grabbed a quick breakfast and arrived at the office at 8 am.  I was very impressed by the accommodations at the office: I was given a desk and two monitors, and could see the One World Trade Center and the Statue of Liberty from my seat on the 42nd floor.  The employees were happy to meet us, and introduced us to the business of the firm and gave us projects to work on.  The idea behind our project was intriguing.  Around noon, our alum, Jerry Hanweck was nice enough to spend an hour of his time describing his firm, its services, and how it fits in with the rest of Wall Street.  This conversation was extremely helpful in allowing us to put into perspective the projects we were working on within the greater world of finance.

On the second day, I arrived at 8:30 am.  Everyone else in the office seemed very busy so we observed for most of the day.  However, there was enough chatter in the office that I could learn through osmosis.  I was able to observe how the office carried out its tasks, how its people collaborated, and how Mr. Hanweck managed his people.

The third day was similar to the first two, except at the end of the day, Mr. Hanweck invited us to his office, and spent an hour talking to us about how he came to starting this business, the firm’s current situation, and its future prospects.  I felt that this was a nice conclusion to the Princeternship.

One of the most valuable parts of the Princeternship was being able to experience daily life on Wall Street.  Having stayed overnight with friends who were bankers, I partook in their day-to-day activities and got a better sense of what my own life would be like if I were to choose this career path.  I think being able to experience the life of a banker outside of the workplace was just as valuable as office work, since office work and culture varies from firm to firm. If it weren’t for the Princeternship program and Mr. Hanweck, I wouldn’t have had the opportunity to experience either.  Thank you!

Wesley Cao ’15, Hanweck Associates, LLC

Fellow Princetern Ryan, Wesley, and Dr. Jerry Hanweck

Day 1

We had just arrived at the office 9 am sharp. The other employees were trickling in, but there was already a buzz around the office. Telephones rang left and right and Bloomberg terminals were under constant surveillance. Mr. Gerry Hanweck ‘87 was on the phone with several clients to see if they had more information.

Apple had just announced a dividend. However, it was unclear what date they set for the dividend end. Apple options will depend heavily on whether Apple goes ex-dividend before or after the expiry date. Hanweck Associates, which specializes in high-computing solutions for option traders, provides an accurate, reliable and most importantly, real-time data stream. Dividends are one of the most important factors in the pricing of options, and undoubtedly they wanted to get the date of Apple’s dividend correct. It could mean a world of difference to the countless option traders in the world.

After this hectic morning, Gerry started us on our first project. As it turns out, it was related to the incident that morning. Hanweck Associates buys analysis from two large firms regarding the ex-dividend dates of various companies. He wanted us to analyze which firm was better at predicting the ex-dividend rate, the yield, and by how much. They then use this data in their own system to better predict the prices of options. It was fascinating to learn what real projects some of their employees might work on. It was also interesting to learn all the factors that went into the pricing of an option. The complexity of the computation process is what led Gerry to start his own firm—by taking advantage of the parallel computing power of graphic processing units, he is able to calculate valuable risk factors and stream them to option traders in real time. We worked on this project for the rest of the day.

Day 2 + 3

We finished up our analysis of dividend forecasts by noon and found some interesting results that we reported to Gerry. It was nothing out of the ordinary and reinforced their existing belief. Then, we promptly started on our second project. Hanweck Associates uses a well-known mathematic model in their computer systems to compute the fair price of options. It currently employs a sophisticated algorithm known as the Fast Fourier Analysis. Gerry wanted us to code another approximation algorithm based on this model. Our algorithm would be simpler, and thus would take less computation speed. This is crucial for Hanweck Associates because they pride themselves on streaming their data in a matter of milliseconds. However, our algorithm would provide less accurate results. Gerry wanted us to investigate exactly how much accuracy is lost through this faster approximation algorithm.

This project seemed very difficult at first; the paper containing the work related to the approximation algorithm was nearly inscrutable. However, with the guidance of another employee, we were able to quickly proceed in our code. This project also allowed me to utilize what I had learned in my classes such as Probability Theory in a practical environment. In the end, we were unable to finish the algorithm completely in the short time that we had, but I still learned a lot of from the experience and am extremely grateful for this wonderful opportunity that Gerry provided us.