Trade against three AI agents in this market making challenge
Instructions
- 1 Make bids and asks on the expected total value of all cards
- 2 Trade against AI players to maximize your profits
- 3 Track your position and manage your risk exposure
- 4 Learn market making principles in a competitive environment
- 5 Analyze card probabilities and make strategic decisions
Market of Cards Explanation
Welcome to “Market of Cards”, a strategic card game that combines market-making and market-taking elements!
Card Values
- Cards 2 to 10 (both red and black) are worth 20 to 100 points, respectively.
- For red suits, Jack, Queen, King, and Ace are worth 110, 120, 130, and 140 points.
- …discover the full scoring system
How to Play
- Each player (you and 3 AI agents) receives 2 cards. 3 cards are placed in the middle.
- Players make markets (bid/ask prices) based on the expected total value of all cards.
- …learn the complete trading mechanics
Strategy Tips
- Pay attention to the revealed cards to adjust your expected values.
- …unlock advanced strategies and techniques
This game is independently developed by Trading Interview B.V. and is protected intellectual property. This proprietary tool is provided for student interview preparation only and may not be reproduced or reverse-engineered.
[16:06:41] You quote: 490 at 510
[16:06:41] AI 1: “Mine at 510”
[16:06:48] AI 1 quotes: 584 at 604
[16:06:54] You: “Yours at 584”
[16:07:26] AI 2 quotes: 429 at 449
[16:07:29] You: “Mine at 449”
[16:07:31] AI 3 quotes: 415 at 435
[16:07:33] You: “Mine at 435”
[16:07:34] My own hand value: 120, Total expected value: 508.5
[16:07:49] You quote: 500 at 520
[16:07:49] AI 1: “Mine at 520”
[16:07:56] AI 1 quotes: 592 at 612
[16:08:00] You: “Yours at 592”
[16:08:03] AI 2 quotes: 437 at 457
[16:08:05] You: “Mine at 457”
[16:08:06] AI 3 quotes: 423 at 443
[16:08:07] You: “Mine at 443”
[16:08:08] My own hand value: 120, Total expected value: 506.2
[16:08:31] You quote: 498 at 518
[16:08:31] AI 1: “Mine at 518”
[16:08:36] AI 1 quotes: 589 at 609
[16:08:39] You: “Yours at 589”
[16:08:40] AI 2 quotes: 435 at 455
[16:08:41] You: “Mine at 455”
[16:08:42] AI 3 quotes: 421 at 441
[16:08:43] You: “Mine at 441”
[16:08:44] My own hand value: 120, Total expected value: 533.8
[16:08:56] You quote: 523 at 543
[16:08:56] AI 3: “Yours at 523”
[16:09:11] AI 1 quotes: 617 at 637
[16:09:13] You: “Yours at 617”
[16:09:14] AI 2 quotes: 462 at 482
[16:09:16] You: “Mine at 482”
[16:09:17] AI 3 quotes: 448 at 468
[16:09:18] You: “Mine at 468”
[16:09:27] It turned out that the total value of the table is 580
[16:09:27]
Your PNL 937 points
AI 1 PNL 130 points
AI 2 PNL -477 points
AI 3 PNL -590 points
How is this PNL even calculated? Simply doing Sells-Buys is not giving the PNL value
Your buys:
Your sells:
So, 9 buys, 7 sells. Cross out 9 buys versus 7 sells to get a PNL of:
The remaining two buys are at 482 and 468, while the table value turned out to be 580:
So, the final PNL for you is 727 + 210 = 937.
This is exactly what the log tells as well, so the game calculated the PNL correctly. Does the calculation in this reply help? If not, don’t hesitate to reach out.
Are we supposed to do these calculations simultaneously while making the market? There are a lot of calculations involved to do simultaneously. Or can we just give some approximate answers during the interviews?
1) I do not understand why you cannot break even for the facts case
2) For the Guesstimate, I got the following result
Pos size + 2
Max possible loss 13,480,000
Your break even is incorrect, the correct answer is sell 2 shares for the price of 8,740,000 each.
Shouldn’t it be 6,740,000?
It shouldn’t be 6,740,000, because you need to add the 6,740,000 on top of the average price of your open two positions (as you mentioned, pos size +2). If after crossing out your lowest sells against your highest buys (to calculate the max guaranteed loss) your remaining positions are, for example, long at 3 million and long at 1 million, your average price being long is 2 million, so you add that on top of the 6,740,000.
Example:
buy at 10 mil
sell at 1 mil
buy at 8 mil
sell at 2 mil
buy at 6 mil
buy at 4 mil
Proof:
Buys:
Sell:
Hope this helps, let me know if anything is unclear.
For calculating the max guaranteed loss, are you supposed to remember all the prices at which you bought & sold, or are you allowed to write things down?
I had the following game:
Please make me a next market with a maximum spread of 5.
I will buy
Please make me a next market with a maximum spread of 5.
I will buy
Please make me a next market with a maximum spread of 5.
I will buy.
Now I will quote the next market. I want to buy 4 lots at 605. Do you want to trade?
Yes, I will sell 4 lots to you (me)
.Please make me a next market with a maximum spread of 5.
I will buy.
Please make me a next market with a maximum spread of 5.
I will sell.
This was the final round, now I will ask a few questions about your position.What is your position? (Example: if you are 2 short, write down -2. If you are 3 long, just write down 3)Position
Your position is incorrect, the correct answer is 1
If I sold 3 times, then sold again 4 times (since it asked to buy from me 4 lots), then sold again, and then it bough from me only 1 time, how can my total position be 1?
Your own EV: no, this is just the EV as it is. It’s up to you if you use it as a mid-market or take on a different strategy.
Your opponents mid market: yes, they take into account their own real-EV, but also include what others quote as well to adjust it.
When I’m done, I’ll confirm it here. Please let me know if you have any questions left afterwards.
buy at 200
buy at 240
sell at 150
sell at 350
The PnL per trade is:
300-200 = 100
300-240 = 60
150-300 = -150
350-300 = 50
Total PnL = 100 + 60 -150 + 50 = 60
10*(1+2+3+4+5+6+7+8+9+10+0+0+0)/13 = 42.3.
-50 + 9*42.3 = 330.8
Important to note: this off course doesn’t take into account that two cards are now out of the deck, and so the expected value per card should have slightly changed. However, the EV just serves as a rough estimate of where your mid-market could be. It would take too much time to constantly calculate the EV too accurately, considering the fact that those exact decimals won’t change your game play. In real life, you won’t have time to calculate it that precisely: the method above is the quickest/ most efficient.