Trading & Quantitative Research

Interview Preparation
Optional reading on trading fundamentals

2024
Trading & Quantitative Research Interview Preparation

Table of Contents

1 Introduction 3
2 Who are we looking for? 4
3 Quantitative trading versus Quantitative research 6
4 Trading basics 7
4.1 Market making 7
4.2 Risks 9
4.3 Position taking 10
4.4 Fermi market making 10
4.5 Markets and confidence 12
5 Conclusion 15
Trading & Quantitative Research Interview Preparation

1 Introduction
At Optiver, our mission is to improve the markets. We believe that an efficient, fair, transparent and
competitive marketplace improves the health of the global economy. How do we accomplish that?
We improve the overall markets by trading a variety of financial instruments using cutting-edge
technology, data-driven strategies and risk-conscious decision making. Optiver is the largest
options market maker in the world, but that is just one of the instrument classes we trade. Given
the trading industry is still relatively small when compared to other careers, it can appear daunting
to those without familiarity. This document’s purpose is to help de-mystify some aspects of the job.

Note that none of this is required reading to successfully pass Optiver’s interview process. We will
challenge your mathematical problem solving ability and decision making, but not in a way that
requires previous trading experience or knowledge of the financial industry. In fact, reading this
document is entirely optional. We only wish to provide an introduction to the world of trading (along
with some concepts and terminology), and perhaps dispel a few interview myths you may have
heard along the way.

3
Trading & Quantitative Research Interview Preparation

2 Who are we looking for?

It is a common misconception that trading requires a specific background. In fact, you don’t need to
have any knowledge of financial math to be successful in the interview. Our applicants come from
a variety of STEM fields: math, engineering, actuarial science, computer science, etc. and our
bespoke training program will teach you everything you need to know about product specifications,
financial instrument pricing and more.

All you need is a sharp, quantitative brain, a love of solving problems and a desire to apply all of
this toward improving the world’s financial markets.

Bring along your curiosity. We will teach you the rest.

What sort of math do you need to know? Given that trading involves risk and reward, probability
and statistics are key. The following three exercises are designed to give you a taste of the types
of things we would like you to be able to think about.

Exercise 1
Two cards are pulled out from a standard deck of 52 cards. What is the probability that they have
the same suit?

Exercise 2
A six-sided dice is rolled six times. What is the probability that the number six does not appear at
all?

Exercise 3
Consider 100 cars all of which are traveling along a road in a randomly chosen order. At first, the
cars are all traveling at distinct speeds and are sufficiently spaced out. Over time, however, some
of the cars will catch up to slower ones and will also have faster cars catch up to them. When this
happens, the faster car that comes up from behind must slow down. Once the cars have all settled
into their clusters, how many of these clusters do you expect there to be?

Exercise 4
You and a friend are going to play a game with a $.50 coin, a $1 coin and $2 coin. You may
choose one coin to be your coin. Then, all three coins will get tossed in the air. If a coin lands on
tails, it counts zero for its owner. If a coin lands on heads, then it counts for its value in cents for
the owner. Whoever gets the largest score wins all three coins (in the case of three tails, the toss is
repeated). What coin should you choose to maximize your expected monetary win?

You do not have to be able to solve these four problems, but you should definitely feel some sort of
internal pull to give them a try! This is what we look for – a drive to solve quantitative problems.

4
Trading & Quantitative Research Interview Preparation

This could be on a computer or using pen or paper.

This also turns out to be one of the best parts about working at Optiver. Our company is full of
brilliant people from different paths who have come together to work on one problem: improving
the world’s financial markets. We do this by providing stable, reliable and accurate pricing to the
market. One mechanism you may have heard of is market making (we will tell you a bit more about
this soon!)

5
Trading & Quantitative Research Interview Preparation

3 Quantitative trading versus Quantitative research

Many of our applicants ask, “Am I more suited to being a Quantitative Trader or a Quantitative
Researcher?” Both roles require a penchant for mathematical thinking and problem solving; it’s the
activities and workflow that are different.

Quantitative Traders spend a good deal of their time plugged in to markets, making decisions
quickly and discussing these decisions in an active environment with colleagues. They write code
and build tools to automate the decision making where possible. Traders are sharp, highly intuitive
and quick at engineering non-textbook approaches to new problems. They are able to influence the
trades we make and the positions we take by controlling the parameters of our trading systems
based on market circumstances. Our Quantitative Traders work in our Chicago and New York
offices and trade on the options desks.

Quantitative Researchers design and deploy algorithms that are meant to operate with a higher
degree of autonomy. While they still monitor these algorithms as they trade in the market, these
employees generate value primarily through their project work and the iterative improvements they
make to these algorithms. This project work makes up the vast majority of their time spent. They
use machine learning, statistical modeling and their domain knowledge to generate features, test
hypotheses and forecast future prices and develop trading strategies around them to profit. While
we accept PhD Quantitative Research candidates in Austin, Chicago, and New York, the
Quantitative Research role for MS/BS candidates is located in our Austin office.

The objectives of the two functions are far more similar than different. Both require curious problem
solvers that design innovative solutions to complicated mathematical problems. Some domains are
friendlier to AI, machine learning and complete automation (think: programming a chess engine).

Other domains are friendlier to a partnership of technology and human decision making (think of
piloting an airplane). So, what role do you think fits you best? Hopefully, the above has helped to
push you in the right direction. You shouldn’t stress, however; employees at Optiver are able to
move between roles if they are looking for a better fit or a new challenge.

Despite the difference in the day-to-day, Quantitative Traders and Quantitative Researchers are
both working on the problem of improving the markets. Let’s take a deeper look at how that might
happen.

6
Trading & Quantitative Research Interview Preparation

4 Trading basics

4.1 Market making

We all make trades every day. Buying a coffee is a trade. Selling the exercise bike you bought
yourself for Christmas but never used is a trade. Waking up early to watch a game of football is a
trade (you are trading sleep for entertainment).

At Optiver, we trade financial instruments. More importantly, in many cases we act as a market
maker: we provide both buy and sell prices for the market participants.

Let’s pretend that there is some instrument out there called X. We can plug ourselves into the
exchange where X is traded and have a peek at the market. It might look something like this:

This is called an order-book. In the middle column, there is a fixed list of prices that X can trade at.
These are listed in descending order.

On the left of each price, you can see a number denoting how many units (also called lots) of X are
desired by buyers at that price. On the right of each price, you can see the number of X that is
wanting to be sold at that price.

So, there could be 17 different people that want to buy one lot of X for $11. Or it might be one
single person who wants to buy 17 lots of X. Or it might be something in between these scenarios.

There are four lots of X waiting to be sold at $14. Again, this might be two different participants out
there. For example, there could be one person wanting to sell one lot and another person wanting
to sell three lots.
7
Trading & Quantitative Research Interview Preparation

Exercise 1
Consider the order-book above. If you had to choose a single price for everyone to do their trades,
what would you choose?

So, in short, there is a marketplace for X consisting of people who want to buy and people who
want to sell. The problem? Right now, as it stands, if a buyer wants to buy then they need to pay
$14. If a seller wants to sell, they will need to sell at $11.

If these trades actually take place, then some amount of X trades at $14 and some amount of X
trades at $11. These are quite large swings in price. In fact, seeing these quick price changes
might even deter some participants from trading as large price swings often indicate a volatile
asset.

Additionally, what if somebody wanted to buy eight lots of X? They would have to pay up through
multiple levels, buying four for $14, two for $15, and the last two for $16.

This is precisely a situation where a market maker is needed. This market is currently illiquid.
Liquidity is a term describing how easy or difficult it is to trade a large size without affecting the
price. In this example, it’s quite difficult for anyone to trade for significant size without dramatically
affecting the price. A trader at Optiver therefore might decide to provide some liquidity. By
combining historical analysis, market signals, current events and any other relevant signals, our
trader works out that their fair theoretical value for X is actually about $12.50. Theoretical value
represents what you believe an asset to be worth.

Our trader then makes a market by inserting both buy and sell prices into the exchange.

Making a market involves providing both a buy and a sell

price for an instrument

They insert a buy order (also called a bid) for 10 lots of X for $12 and a sell order (also called an
offer) for 10 lots of X at $13. Note that neither of these orders results in a trade – yet. They have
indicated to the market that they would be willing to buy X for $12 and would also be willing to sell
X for $13.

Our trader should be pretty happy if they trade either order. They believe that X is worth $12.50
and is therefore happy to buy it at $12 or sell it for $13. Indeed, the trader would expect to make a
theoretical profit of $.50 per lot that they trade.

8
Trading & Quantitative Research Interview Preparation

The market participants are happy here too. This market is now more liquid. A buyer can buy it
immediately for $13 and a seller can sell it immediately for $12. The market is now a fairer place for
its participants.

Notably, as the participants trade, the market maker generates profits as they buy at $12 and sell
at $13 repeatedly.

4.2 Risks
There are a few risks that might spring to mind from this simple market making example.

What if a buyer immediately buys the 10 lots our trader is offering at $13? Our trader will then have
a short position in the instrument X. If X increases in value, our trader will be losing money. Don’t
forget that markets can move – and quickly, too!

Exercise 2
If our trader sells 10 lots at $13, then the market moves so that they now believe the fair price for X
is $17, what is their theoretical P/L (profit/loss)?

Profit or loss is calculated by comparing the current price to the traded price, then multiplying by
volume. When you buy, you profit from a rise in the price of the asset. Thus your P/L is:

(Current Price – Purchase Price) * Volume

When you sell, your profit from a decline in the price of the asset. Thus your P/L is:

(Sell Price – Current Price) * Volume

In our example, our trader would have lost $40. It’s not that unlikely that you might run into
scenarios like this where you find it easier to sell X but hard to buy, or vice versa. Go back and
9
Trading & Quantitative Research Interview Preparation

have a glance at the market for X. Something observable is at play here – demand. There actually
does seem to be more buyers than sellers in this market. This can create upward pressure on X. A
good mathematical model should likely take this into account.

Exercise 3
What other risks are involved in market making?

4.3 Position taking

After reading the previous section, it might appear that the objective of a market maker is therefore
to stay flat. After all, buying and selling an equal amount of instrument X would leave our trader
with no leftover risk and only profit. However, it is a misconception that a market maker’s objective
is to always remain flat.

Exercise 4
What if our trader thought that the fair theoretical value for X is $17? What should their market be?

Sometimes, you might encounter a scenario like this where you disagree with others in the market
about what the fair value should be. Trading involves predicting the future, and different market
participants use different models, information and features to create predictions. If your theoretical
value is quite different from other people, it can be tricky to navigate. This is not all of the possible
tactics you could use, but here you might consider:

 Buying X for $14 from the participants willing to sell there

 Inserting a bid for $13 instead of $12
 Inserting a bid for $12, but not inserting an offer at $13

Each of these routes has its own advantages and disadvantages. Thinking through the best
strategy is often one of the most challenging parts of trading. In all cases though, your objective is
not to buy and sell equal amounts of X, but just to buy because you believe the price of X will rise.
When you actively seek to buy or sell without trying to do an offsetting trade, we call that position
taking.

4.4 Fermi market making

Optiver’s focus is on improving financial markets, but good practice can actually be developed by
making markets on the solutions to Fermi problems.

These are named after the physicist Enrico Fermi, who was quite fond of estimating unknown
quantities with little information. The most striking example of this is the time he dropped some

10
Trading & Quantitative Research Interview Preparation

paper at a (safe) distance from an atomic blast and used the distance the paper traveled (quite
accurately!) to estimate the energy released from the bomb.

Have a go at the following Fermi problem as a warm-up.

Exercise 5
How many people are on an airplane in the sky right now?

In this section, we will consider the act of making a market on a Fermi problem. Let’s say that a
friend of yours taps you on the shoulder and gives you the following seemingly strange request:

Make me a market on the perimeter of Tasmania.

Let’s unpack this a bit. The perimeter of Tasmania is the total length of the boundary. Your friend
wants you to provide a buy and sell price on this length.

A more thorough explanation might be:

I want you to give me two numbers B and S. The first number B is your buy price, and this should
be below what you think the perimeter of Tasmania is. The second number S is your sell price, and
this should be above what you think the perimeter of Tasmania is.

OK, it’s getting clearer. Your friend is trying to get you to come up with an estimate of this distance,
but she doesn’t want your estimate. She just wants two numbers on either side of it.

You think for a bit. After a while you come up with an estimate of about 1200KM. You then jump
400KM each way from your estimate and tell your friend that you have chosen B = 800KM and S =
1600KM.

Congratulations – you have just made a market. If your friend thinks that the true answer is less
than 800KM, then she will sell you 800KM. After all, you have shown her a buy price there, so she
is well within her rights to do this! On the other hand, if she thinks the answer is greater than
1600KM, then she will buy this from you.

Your friend opens her mouth and says:

I will buy 1600KM from you.

You and your friend have just done a trade! She has bought “the perimeter of Tasmania” from you
at the price of 1600KM.

11
Trading & Quantitative Research Interview Preparation

Note that you should both be happy with your trades. You have sold at a price of 1600KM when
you believe the answer to be around 1200KM. Your friend has bought at a price of 1600KM when
she believes the answer to be greater than 1600KM.

Of course, only one of you can be correct. Let’s proceed to settlement; this is where the solution to
our Fermi problem is revealed. In this case, a quick check on Google gives the (approximate)
answer of 1400KM.

You sold at a price of 1600KM and the thing you sold settled at 1400KM. Therefore, you have
made a profit of 200KM. Your friend has made a loss of 200KM.

Note: It certainly seems odd to talk about profits in terms of kilometers. If you and your friend had
done this properly, you likely would have agreed upon some conversion rate (example: $1 per
10KM).

4.5 Markets and confidence

Suppose I ask you to make me markets on the following:

 The world record for the longest distance (in meters) a golf ball has been thrown
 The total number of words in all seven of the (English) Harry Potter books
 The probability that two randomly chosen positive integers have no factors in common

In each case, you first need to come up with an estimate. And not only that, but your estimates
must be sane. If you say that the world record for the golf ball throw is 2KM or you say that the total
number of words in all seven of the Harry Potter books is 400, then something is clearly wrong.

Once you come up with an estimate or fair value, you need to provide your buy price, B, and sell
price, S. But how to pick these?

If you think that the world record for the golf ball throw is 150M, then here are two possible markets
you might provide:

Market 1: B = 0M S = 300M
Market 2: B = 149M S = 151M

Neither of these is great. Consider walking into a room (or the Optiver trading floor) where
everyone wants to trade on the world record for the longest distance a golf ball has been thrown. In
fact, they have been waiting for a market maker like you to provide them with liquidity!

If you announce Market 1 to everybody, you will likely do zero trades. Nobody out there wants to
sell 0M. Maybe a couple of people will buy 300M. But really, your market is too wide, and as such
does not provide any real liquidity.

12
Trading & Quantitative Research Interview Preparation

On the other hand, if you announce Market 2, you will probably do a lot of trades! If someone
thinks that the answer is smaller than 149M or larger than 151M then they will trade with you. This
is going to encapsulate basically everyone in the room. So, well done: you’ve provided some real
liquidity! The problem is risk.

The ideal situation with Market 2 is that half the people in the room sell you 149M and the other
half of the room buy 151M. You neatly offset every buy with a sell and you generate profit
(assuming you agreed on a settlement conversion rate). Every buy that you have done at 149m is
neatly covered with a sell that you have done at 151M. You make a profit of 2M (the width of your
market) for every buy-sell pair.

However, as your (B, S) interval is so tight, it is much more likely that the consensus of the room
falls to the side of your interval. To put some numbers to it, let’s say that there are 10 people in the
room and that four of them sell to you and the other six buy from you.

Your four buys at 149M pair up nicely with four of your sells at 151M. These eight trades get you a
profit of 4 × 2m = 8m. But there are two more sells that you have done at 151M that you have not
covered on the buy side.

If the answer turns out to be 180M, then you will lose 29M on each of these sells at 151M. These
losses massively override your profits and your golf-ball market making career is sadly over.

The above example demonstrates why Market 2 is not great. The amount that you are getting paid
to trade (a function of your width) is too small relative to the uncertainty in the underlying value.

The width of your market should be related to the

uncertainty of the thing you are trading

Sometimes, we have information, and this should change our market. Also, perhaps we are
constrained in some way.

Think about different information and constraints and how these can affect the width and position of
your market.

To end this section, have a think about the following scenario.

Exercise 6
What if you had walked into the trading room late in the proceedings and watched for a little bit?

13
Trading & Quantitative Research Interview Preparation

Let’s say you just saw trading happen at these prices, in this order:

149M,152M,160M,155M,150M,149M,153M,158M.

What market would you make?

14
Trading & Quantitative Research Interview Preparation

5 Conclusion

As you can see, trading at its core is quantitative decision making. Your job is to predict the future
and devise a strategy under uncertainty. During the interview, your intellectual curiosity and
approach you take to problem solving are weighed heavily. A sound decision might suffer an
unlucky result, and a poor decision can sometimes get lucky. Whenever possible, we encourage
you to think out loud and talk through your process. We hope you found this valuable. Best of luck!

15