Wednesday, December 10, 2014
New Thoughts!
After what seems to be a while, I am keen to revisit the game in order to simplify the structure to make it into a more educational program. So, thinking six formers studying economics, business studies and, of course, mathematicians.
Team Outline.
Market Maker (MM)- No change there
2x Locals (PL)- No change, but make sure it is the noisiest or the most confident
2x Hedge Fund Mangers (HF)- In this scenario two teachers.
4x Brokers - No investment bankers now, each team has three members - a trader in the pit (PT), a clerk (BC) and a sales-trader (ST).
The big change is the makeup of the pit which is now seven, so hopefully more noisy and not completely dependent on the MM creating the noise. The Roles within the Broker is described as follows:-
ST - negotiates prices and levels with each of the HF clients in size of a minimum of 50 and executes bilaterally with the HF but with nobody else. The ST needs to know from the BC what position they are running and change the quote to the HF to reflect the axe. The ST is also responsible for the settlement of the trade and telling the BC of a position change.
BC - keeps track of the overall position, issues orders to the PT to execute for the broker. The BC also takes whatever trades the PT has completed and settles them. The BC can advise (hopefully with hand signals) the ST on the market level in the pit and overall position. The BC cannot enter the pit.
PT - stands in the pit and execute trades that fill the order from the BC hopefully. Executes with other pit members, so other traders, the MM and PL. The PT can trade in anticipation of an order. The PT does not leave the pit so hands the completed tickets to the BC.
PL - locals trade on their own account and are not obliged to makes a two-way market, only the MM has to do that.
Given there are six rounds and three roles, the idea would be to rotate the members in the Broker after two and four rounds.
Saturday, March 26, 2011
Thoughts post 21 March game
During the game we had, 1 MM, 1 Local, 1 local firm with 2 players, four investment banks with 3 each, two hedge funds, one consumer two producers each with two.
1) An evolution that happened naturally was the locals took limit orders from hedge funds which was a good idea and we will allow that going forward. Upto individual locals if they can cope with the requests. One suggestion is to split the Locals into either taking all buy orders or all sell orders, this reduces the opportunity for confusion
2) We have simplified the spreadsheet that supports the settlement. Please send requests to riskpearlwisdom@gmail.com and I will find a way of returning by email
3) This game we closed 180 deals so having a large number of cards was useful, as ever there is about 30% loss due to errors and mis-markings etc
4) We created name cards for everyone and allocated names rather than allowing people to come up with own - so we had MM, L1, L2, IB1, ... , IB4, HF1, HF2, C1, P1, P2
5) one suggestion is to have two consumers and two producers and suggest that they are in competition so even if you lose money, so long as you lose less than the other consumer then its good
6) the opening after each start was done by open outcry which slowed the game down a bit, need a better solution. maybe we collect the limit orders and use that to open the market. Alternatively we return to original approach of all IB forcing to trade on the open
1) An evolution that happened naturally was the locals took limit orders from hedge funds which was a good idea and we will allow that going forward. Upto individual locals if they can cope with the requests. One suggestion is to split the Locals into either taking all buy orders or all sell orders, this reduces the opportunity for confusion
2) We have simplified the spreadsheet that supports the settlement. Please send requests to riskpearlwisdom@gmail.com and I will find a way of returning by email
3) This game we closed 180 deals so having a large number of cards was useful, as ever there is about 30% loss due to errors and mis-markings etc
4) We created name cards for everyone and allocated names rather than allowing people to come up with own - so we had MM, L1, L2, IB1, ... , IB4, HF1, HF2, C1, P1, P2
5) one suggestion is to have two consumers and two producers and suggest that they are in competition so even if you lose money, so long as you lose less than the other consumer then its good
6) the opening after each start was done by open outcry which slowed the game down a bit, need a better solution. maybe we collect the limit orders and use that to open the market. Alternatively we return to original approach of all IB forcing to trade on the open
Wednesday, March 16, 2011
Feedback from Trading Game 14 March
We had four IB, two HF, one C, one P, two LT and one MM.
This was a group new to the game so we had three in each IB, a two person HF and a one person HF, two in each of C and P and one experienced LT and two in a second LT. Couple of observations:-
1) we tried the news service and it was one step too far, so next time this will be ditched.
2) putting two locals together to work as a team was a good idea, so they serviced one of the C or P and the more experienced local serviced the other. He also gave the new Locals advice on how to trade from the pit. All in all worked out well.
3) the C and P plus the limit orders seemed to work well, each had 200 contracts to unwind and they managed to do it entirely using the limit order process. The locals managed to use the info well and trade around the order. Also charging the C and P one point fee worked well.
4) I still think that the MM should force everyone to trade on the open, though this should be seen as to make sure they can settle trades properly. After the first round, this requirement can be relaxed if there is sufficient trading going on. Forcing all IB to trade with the MM at the start of each round doesnt add. In this game we closed over 130 trades so forcing everyone to trade on open was not required. So after round 1, I would drop it.
5) the use of different coloured cards for buy and sell was a hit. Also allow for people to make a mess of filling them in, we had 200 each buy and sell and well ran out.....hmmm. Next time we will double the order.
6) some quotes from the IB to the HF were too wide - 170/185 was the worst, so set a 5 point max spread.
7) we had some crazy swings in PL. One team went down 6000 before recovery. My feeling is that if at the close of a trading session a team is under water by 2000, they must cut their holding by half by the end of the next session. If they hit a loss of 3000, they need to cut their holding by a further half or to an open position of 50, whichever is smaller. By 5000 down, they need to unwind completely.
8) For locals that are under water by 1000, they trade out immediately with the MM with a five point margin away from latest quote. They can continue until they are down 2000, when the MM will take over their positions and expel them from the pit. This is unlikely to happen if they focus on serving their clients.
9) single HF trader didnt work, he had no chance of keeping track of the market.
10) hand signals worked well in the pit, so did the yellow jackets!
This was a group new to the game so we had three in each IB, a two person HF and a one person HF, two in each of C and P and one experienced LT and two in a second LT. Couple of observations:-
1) we tried the news service and it was one step too far, so next time this will be ditched.
2) putting two locals together to work as a team was a good idea, so they serviced one of the C or P and the more experienced local serviced the other. He also gave the new Locals advice on how to trade from the pit. All in all worked out well.
3) the C and P plus the limit orders seemed to work well, each had 200 contracts to unwind and they managed to do it entirely using the limit order process. The locals managed to use the info well and trade around the order. Also charging the C and P one point fee worked well.
4) I still think that the MM should force everyone to trade on the open, though this should be seen as to make sure they can settle trades properly. After the first round, this requirement can be relaxed if there is sufficient trading going on. Forcing all IB to trade with the MM at the start of each round doesnt add. In this game we closed over 130 trades so forcing everyone to trade on open was not required. So after round 1, I would drop it.
5) the use of different coloured cards for buy and sell was a hit. Also allow for people to make a mess of filling them in, we had 200 each buy and sell and well ran out.....hmmm. Next time we will double the order.
6) some quotes from the IB to the HF were too wide - 170/185 was the worst, so set a 5 point max spread.
7) we had some crazy swings in PL. One team went down 6000 before recovery. My feeling is that if at the close of a trading session a team is under water by 2000, they must cut their holding by half by the end of the next session. If they hit a loss of 3000, they need to cut their holding by a further half or to an open position of 50, whichever is smaller. By 5000 down, they need to unwind completely.
8) For locals that are under water by 1000, they trade out immediately with the MM with a five point margin away from latest quote. They can continue until they are down 2000, when the MM will take over their positions and expel them from the pit. This is unlikely to happen if they focus on serving their clients.
9) single HF trader didnt work, he had no chance of keeping track of the market.
10) hand signals worked well in the pit, so did the yellow jackets!
Saturday, February 19, 2011
Summary of rules for a forthcoming session
I have been asked to run a trading simulation for colleagues as part of a training program. There are 30 participants, so will be introducing Producers and Consumers into the mix, also the Insider who peddles rumours based on News. The teams are split as follows:-
1 Market Marker - this is me
3 Locals
4 Investment Bankers with 4 people in each
2 Hedge Fund Managers with 2 in each
2 Consumers with 2 in each
4 Producers with 1 in each
1 dodgy day trader - colleague
1 Settlement person from the Training team.
1) The purpose of the game is to simulate the functions of four participants in a financial trading market – Market-Makers (M), Locals (L) Investment Bankers (B), Hedge-Fund clients (H), Producers (P) and Consumers (C). For the purpose of this simulation we have 1 Market-maker, 3 Locals who together form the Pit, 4 Investment Banks, 2 Hedge Funds, 4 Producers and 2 Consumers.
2) The game is based on trading in a pit environment that also includes curb or off-pit trading and most importantly settlement. Trading is done face-to-face using paper tickets. These tickets are settled in real time by the Settlement Administrator (S). Once each trade is settled by S, that becomes the prime record and is not open to dispute. Financial penalties will be incurred by those who are unable to execute and settle a trade accurately. For the smooth and quick running of the game, please pay attention to the simple rules of price, size, buy/sell to enable settlement to happen smoothly.
3) The game is zero-sum gain so the primary test for each group is to maximise profitability by executing a trading and hedging strategy. Therefore credit is given to all players for the volume of deals that they execute as well as their PL profile during the game. Those that sit on the sideline will be marked poorly. At this point, each group should choose their team's two letter acronym (e.g. "PH") and write them on their name badges and notify S. This is each team's trading ID and is essential to trading and settlement.
4) All groups are trading contract for differences (CFD) whose final level is based on the sum value of the 14 cards that are chosen from two packs of cards. The first pack of cards is numbered 1 to 40. M shuffles pack 1 and hands out 2 cards to each of B. Only B see their 2 cards and therefore the cards are blind to M, L, H, P and C. The second pack contains 100 cards that have been selected from a standard normal distribution multiplied by 10 – note they can be positive as well as negative but their sum is exactly equal to zero. M shuffles pack 2 and chooses six cards and places them face down in the centre. These cards are blind to everyone and with the 8 cards in pack 1 make up the 14 cards that will define the closing price of the contract.
5) Using a simulation of 100,000 games, the average sum of 8 cards from pack 1 is 164 and standard deviation of 29.5, the average for 6 cards from pack 2 is zero with standard deviation of 25. Together, the sum of 14 cards average is 164 with a standard deviation of 39. For those of a mathematical bent, the distribution is Normal and not LogNormal.
6) The role of M is to simply make a continuous two-way market for L and B to trade with up to a maximum size of £5 a point with a spread of 3. Minimum trade size is £2 a point. Being your host for the next two hours, he has no incentive to make money, only to avoid losing and therefore his function is price discover - find a level where there are as many buyers as sellers. M's main advantage is that he sees the most volume of trades and can manage his position accordingly. M also calls the end of each of the six sessions that usually co-incides with achieving a flat position.
7) M is joined in the pit by L who are locals that trade on their own account, nobody else is allowed in the pit. They do not have the obligation that M has to make a continuous two-way market and can show prices inside M spread in size when they feel fit. Their function is to provide liquidity when M is struggling to flatten out and are expected to make some noise while in the pit. They are required to service their clients who are P and C by executing limit orders. Locals have limited capital and are not allowed to carry excessive losses. If S notices that at the end of a session, a L has lost more than £1000 based on mid price shown by M, S can force the liquidation of L and trade the position out with M at three points outside of the last quoted spread. At this point L has to exit the pit and seek a job with one of the other groups.
8) B have multiple functions and therefore will have the largest teams. Firstly, B uses their proprietary information based on their two cards from pack 2 to trade with M and L and take a directional view. Secondly, B must service each H by showing a sealed two-way price in size of at least £20 a point and no more than a 5 point spread. They do this at least once a trading session for each H. B will be judged by how much they generate volume of trading from H clients. Thirdly, B can service P and C by showing them prices when asked or can be more aggressive and show axes to P and C. There is no obligation to respond to requests or show them competitive prices but again B should find ways to attract as much volume as possible and B should benefit from additional proprietary information.
9) H receive private/sealed prices from each of the B during each trading period and are expected to take sizeable directional views. The private quotes present arbitrage opportunities as well as proprietary information on how the B are positioned which H must try to exploit. The only and key condition is that they H are required to have a flat position at the end of the game. They do not have access to anyone other player than B for liquidity, though M will get involved in the last session to help to close out open positions...and it wont be cheap for H.
10) The game simulates the physical component of financial trading, so P and C are the initiators and end-consumers of the underlying physical that the CFD is based on. P and C start the game by execution large trades with M to represent their natural exposure to the physical. P, being producers benefit from an increase in price and C benefit from a reduction. Therefore P execute a Buy trade with M on the opening spread and C execute a Sell with M. Their strategy is to reduce their position to within £20 a point without losing money. After the initial trade, P and C can only access the Pit by leaving limit orders with L who try to fill the order during a single session. Any unfilled orders expire at the end of each session. P and C can place up to two orders with any L. P and C can also approach any B for a one-way quote and can choose to execute or not, thought B can refuse to show prices if they are deemed time wasters. P and C cannot trade directly with H or M.
11) The game starts with six cards from pack 2 turned down and M opens the first session with a 3-point spread in up to £5 a point. To open the trading, M executes opening positions with each P and C on the spread. This is the only real income that M receives. After that, every B must trade on this open quote - this facilitates price discovery and forces everyone to then have a trading axe. Only one person from each B is allowed to approach the pit at any one time, otherwise it gets crowded. Once each P, C and B has traded with M, the market is now open to L who can join in showing alternative prices and receive orders from P and C. At this point M will make subsequent prices to balance his book. Each B must now show their private and fixed prices to each H who must give them fills quickly where they want to. The session will continue until trading dries up and M calls the end bell. At that point all trading stops, all trade settlement is completed and at that point M will turn over one of the six cards.
12) Session 2 through to six continue as session one with M opening prices and every B trading on the new spread. After M has turned over the last card in pack 2, the game enters the final session where the rule that H have to close out their positions is imposed and that each of P and C have reduced their position to below £20. M also has to finish the game with a flat position. B and L are allowed to retain holdings as they see fit. Once M signals the end of the last session, the eight cards from the B are revealed and M will calculate the closing level for the contract. All outstanding contract then settle on that level.
13) Settlement Rules. B can trade with M or L, however once they have completed a trade, they then go IMMEDIATELY to S and present two tickets - one for Buy and one for Sell with the same deal size and price. When executing amongst M and L, L takes the two tickets IMMEDIATELY to S. When B trade with H, P or C then each B take the two tickets IMMEDIATELY to S. P and C fill in limit order forms and give them directly to L. When L completes the order, a buy and sell ticket is created between L and the P or C take the two tickets IMMEDIATELY to S. All deals that L has executed to fill the order are settled with L as principle.
14) Once S receives two tickets, the trades are recorded in the settlement spreadsheet. If the tickets match then the representation in the spreadsheet is now the prime record and no further disputes are allowed. S also keeps track of latest quotes from M to keep a check on each groups PL. Groups should also not crowd S - they are allowed to check their positions but are not allowed to see others. S sits at the back of the pit behind M.
15) There is a fifteenth card and a dodgy day trader! During each session a rumour about news that will have an impact on the market of either -10, -5, 0 (for silly rumours), +5 and +10. Each group should look out for the dodgy day trader (D) person in the room with early news knowledge and decide to work with D and act on whatever information there is to share. He can execute with anyone outside of the Pit in small size up to £5 a point as a way of sharing the gains that come from the information. Each news item is flashed at the end of each trading session. The sum of the impact of the news items is held in the fifteenth card and will be unveiled by M and added to the original 14 cards. Nobody including D knows the content of the fifteenth card, that will be provided by the Training team.
1 Market Marker - this is me
3 Locals
4 Investment Bankers with 4 people in each
2 Hedge Fund Managers with 2 in each
2 Consumers with 2 in each
4 Producers with 1 in each
1 dodgy day trader - colleague
1 Settlement person from the Training team.
1) The purpose of the game is to simulate the functions of four participants in a financial trading market – Market-Makers (M), Locals (L) Investment Bankers (B), Hedge-Fund clients (H), Producers (P) and Consumers (C). For the purpose of this simulation we have 1 Market-maker, 3 Locals who together form the Pit, 4 Investment Banks, 2 Hedge Funds, 4 Producers and 2 Consumers.
2) The game is based on trading in a pit environment that also includes curb or off-pit trading and most importantly settlement. Trading is done face-to-face using paper tickets. These tickets are settled in real time by the Settlement Administrator (S). Once each trade is settled by S, that becomes the prime record and is not open to dispute. Financial penalties will be incurred by those who are unable to execute and settle a trade accurately. For the smooth and quick running of the game, please pay attention to the simple rules of price, size, buy/sell to enable settlement to happen smoothly.
3) The game is zero-sum gain so the primary test for each group is to maximise profitability by executing a trading and hedging strategy. Therefore credit is given to all players for the volume of deals that they execute as well as their PL profile during the game. Those that sit on the sideline will be marked poorly. At this point, each group should choose their team's two letter acronym (e.g. "PH") and write them on their name badges and notify S. This is each team's trading ID and is essential to trading and settlement.
4) All groups are trading contract for differences (CFD) whose final level is based on the sum value of the 14 cards that are chosen from two packs of cards. The first pack of cards is numbered 1 to 40. M shuffles pack 1 and hands out 2 cards to each of B. Only B see their 2 cards and therefore the cards are blind to M, L, H, P and C. The second pack contains 100 cards that have been selected from a standard normal distribution multiplied by 10 – note they can be positive as well as negative but their sum is exactly equal to zero. M shuffles pack 2 and chooses six cards and places them face down in the centre. These cards are blind to everyone and with the 8 cards in pack 1 make up the 14 cards that will define the closing price of the contract.
5) Using a simulation of 100,000 games, the average sum of 8 cards from pack 1 is 164 and standard deviation of 29.5, the average for 6 cards from pack 2 is zero with standard deviation of 25. Together, the sum of 14 cards average is 164 with a standard deviation of 39. For those of a mathematical bent, the distribution is Normal and not LogNormal.
6) The role of M is to simply make a continuous two-way market for L and B to trade with up to a maximum size of £5 a point with a spread of 3. Minimum trade size is £2 a point. Being your host for the next two hours, he has no incentive to make money, only to avoid losing and therefore his function is price discover - find a level where there are as many buyers as sellers. M's main advantage is that he sees the most volume of trades and can manage his position accordingly. M also calls the end of each of the six sessions that usually co-incides with achieving a flat position.
7) M is joined in the pit by L who are locals that trade on their own account, nobody else is allowed in the pit. They do not have the obligation that M has to make a continuous two-way market and can show prices inside M spread in size when they feel fit. Their function is to provide liquidity when M is struggling to flatten out and are expected to make some noise while in the pit. They are required to service their clients who are P and C by executing limit orders. Locals have limited capital and are not allowed to carry excessive losses. If S notices that at the end of a session, a L has lost more than £1000 based on mid price shown by M, S can force the liquidation of L and trade the position out with M at three points outside of the last quoted spread. At this point L has to exit the pit and seek a job with one of the other groups.
8) B have multiple functions and therefore will have the largest teams. Firstly, B uses their proprietary information based on their two cards from pack 2 to trade with M and L and take a directional view. Secondly, B must service each H by showing a sealed two-way price in size of at least £20 a point and no more than a 5 point spread. They do this at least once a trading session for each H. B will be judged by how much they generate volume of trading from H clients. Thirdly, B can service P and C by showing them prices when asked or can be more aggressive and show axes to P and C. There is no obligation to respond to requests or show them competitive prices but again B should find ways to attract as much volume as possible and B should benefit from additional proprietary information.
9) H receive private/sealed prices from each of the B during each trading period and are expected to take sizeable directional views. The private quotes present arbitrage opportunities as well as proprietary information on how the B are positioned which H must try to exploit. The only and key condition is that they H are required to have a flat position at the end of the game. They do not have access to anyone other player than B for liquidity, though M will get involved in the last session to help to close out open positions...and it wont be cheap for H.
10) The game simulates the physical component of financial trading, so P and C are the initiators and end-consumers of the underlying physical that the CFD is based on. P and C start the game by execution large trades with M to represent their natural exposure to the physical. P, being producers benefit from an increase in price and C benefit from a reduction. Therefore P execute a Buy trade with M on the opening spread and C execute a Sell with M. Their strategy is to reduce their position to within £20 a point without losing money. After the initial trade, P and C can only access the Pit by leaving limit orders with L who try to fill the order during a single session. Any unfilled orders expire at the end of each session. P and C can place up to two orders with any L. P and C can also approach any B for a one-way quote and can choose to execute or not, thought B can refuse to show prices if they are deemed time wasters. P and C cannot trade directly with H or M.
11) The game starts with six cards from pack 2 turned down and M opens the first session with a 3-point spread in up to £5 a point. To open the trading, M executes opening positions with each P and C on the spread. This is the only real income that M receives. After that, every B must trade on this open quote - this facilitates price discovery and forces everyone to then have a trading axe. Only one person from each B is allowed to approach the pit at any one time, otherwise it gets crowded. Once each P, C and B has traded with M, the market is now open to L who can join in showing alternative prices and receive orders from P and C. At this point M will make subsequent prices to balance his book. Each B must now show their private and fixed prices to each H who must give them fills quickly where they want to. The session will continue until trading dries up and M calls the end bell. At that point all trading stops, all trade settlement is completed and at that point M will turn over one of the six cards.
12) Session 2 through to six continue as session one with M opening prices and every B trading on the new spread. After M has turned over the last card in pack 2, the game enters the final session where the rule that H have to close out their positions is imposed and that each of P and C have reduced their position to below £20. M also has to finish the game with a flat position. B and L are allowed to retain holdings as they see fit. Once M signals the end of the last session, the eight cards from the B are revealed and M will calculate the closing level for the contract. All outstanding contract then settle on that level.
13) Settlement Rules. B can trade with M or L, however once they have completed a trade, they then go IMMEDIATELY to S and present two tickets - one for Buy and one for Sell with the same deal size and price. When executing amongst M and L, L takes the two tickets IMMEDIATELY to S. When B trade with H, P or C then each B take the two tickets IMMEDIATELY to S. P and C fill in limit order forms and give them directly to L. When L completes the order, a buy and sell ticket is created between L and the P or C take the two tickets IMMEDIATELY to S. All deals that L has executed to fill the order are settled with L as principle.
14) Once S receives two tickets, the trades are recorded in the settlement spreadsheet. If the tickets match then the representation in the spreadsheet is now the prime record and no further disputes are allowed. S also keeps track of latest quotes from M to keep a check on each groups PL. Groups should also not crowd S - they are allowed to check their positions but are not allowed to see others. S sits at the back of the pit behind M.
15) There is a fifteenth card and a dodgy day trader! During each session a rumour about news that will have an impact on the market of either -10, -5, 0 (for silly rumours), +5 and +10. Each group should look out for the dodgy day trader (D) person in the room with early news knowledge and decide to work with D and act on whatever information there is to share. He can execute with anyone outside of the Pit in small size up to £5 a point as a way of sharing the gains that come from the information. Each news item is flashed at the end of each trading session. The sum of the impact of the news items is held in the fifteenth card and will be unveiled by M and added to the original 14 cards. Nobody including D knows the content of the fifteenth card, that will be provided by the Training team.
Wednesday, February 16, 2011
Producers and Consumers
I have been asked to extend the number of participants to 30 instead of the original design. This gives me the opportunity to introduce a new set of players which complete the list of participants in a trading environment. I take the cue from commodities that have producers and consumers (PR, CO for short).
Each PR and CO start with an open position which is marked at the beginning either at the theoretical price or at the mid of the opening price shown by the MM, doesnt really matter. The size of the PR and CO should be the same, but if you want to have more than one PR and CO, you should have more PRs than CO. Key is that the sum of all positions should be net zero but large from a gross point of view.
PRs are naturally long as they benefit from a bull market, so they start with a long position. COs are naturally short as their costs increase in a bull market, so they start with a short position.
The motivation for each PR and CO is to break even or make money by winding down their exposure. Their decision making is one of timing and risk management - how much can they offload and when is the best time. By the end of the game they must hedge at least 80% of their holding. They can attempt a squeeze on the market by increasing their exposure (i.e. PR increasing their long position at the outset) before unwinding their view.
In terms of trading, they do not have access to the open market - the one created by the MM or by LT. They also cant talk to the HF. They have access to the pit via the LT and the mechanism is limit orders.
A limit order sets a size and level by which the LT can execute in the pit, the order is valid for a single trading session. The order is given on a bi-lateral basis between the PR/CO and their LT so is not known by anyone else. The LT must build a relationship with the PR/CO so that they feel that they are working for their advantage. It is up to the LT to decide when to execute and in what size - the LT doesnt have to complete the fill but at the end of the day they want to increase their reputation so are allowed to cross the order with themselves.
The limit order gives them the optionality to quicky cross the trade in a moving market and profit from any further move. This legging over is entirely illegal in normal markets but is a fun way to generate some tension. The LT can always hand over a better execution if his conscience dictates!
The PR can give upto 2 limit orders, say 10 at 180 and 20 at 185, but no more.
The PR and CO can contact the IB for immediate access to liquidity. The IB can make any price they like and in any size they want but should be a worse price than what is available by the MM. IF a PR or CO does ask for a price from an IB and they dont like the price, they can refuse. There is a lot of opportunity of messing around here which is what happens in reality, if a PR is using the IBs for price discovery and never trades, then the IBs can simply refuse further requests and the client is frozen out.
The size of the initial positions in each PR and CO should be material, say if we have 2 PR and 1 CO, I would put them at +100 position for each PR and -200 for the COs.
Each PR and CO start with an open position which is marked at the beginning either at the theoretical price or at the mid of the opening price shown by the MM, doesnt really matter. The size of the PR and CO should be the same, but if you want to have more than one PR and CO, you should have more PRs than CO. Key is that the sum of all positions should be net zero but large from a gross point of view.
PRs are naturally long as they benefit from a bull market, so they start with a long position. COs are naturally short as their costs increase in a bull market, so they start with a short position.
The motivation for each PR and CO is to break even or make money by winding down their exposure. Their decision making is one of timing and risk management - how much can they offload and when is the best time. By the end of the game they must hedge at least 80% of their holding. They can attempt a squeeze on the market by increasing their exposure (i.e. PR increasing their long position at the outset) before unwinding their view.
In terms of trading, they do not have access to the open market - the one created by the MM or by LT. They also cant talk to the HF. They have access to the pit via the LT and the mechanism is limit orders.
A limit order sets a size and level by which the LT can execute in the pit, the order is valid for a single trading session. The order is given on a bi-lateral basis between the PR/CO and their LT so is not known by anyone else. The LT must build a relationship with the PR/CO so that they feel that they are working for their advantage. It is up to the LT to decide when to execute and in what size - the LT doesnt have to complete the fill but at the end of the day they want to increase their reputation so are allowed to cross the order with themselves.
The limit order gives them the optionality to quicky cross the trade in a moving market and profit from any further move. This legging over is entirely illegal in normal markets but is a fun way to generate some tension. The LT can always hand over a better execution if his conscience dictates!
The PR can give upto 2 limit orders, say 10 at 180 and 20 at 185, but no more.
The PR and CO can contact the IB for immediate access to liquidity. The IB can make any price they like and in any size they want but should be a worse price than what is available by the MM. IF a PR or CO does ask for a price from an IB and they dont like the price, they can refuse. There is a lot of opportunity of messing around here which is what happens in reality, if a PR is using the IBs for price discovery and never trades, then the IBs can simply refuse further requests and the client is frozen out.
The size of the initial positions in each PR and CO should be material, say if we have 2 PR and 1 CO, I would put them at +100 position for each PR and -200 for the COs.
Wednesday, February 2, 2011
Introducing News Shocks
A couple of people (Priyesh and Gary) questioned why there is not a process by which shocks to the market can be introduced to the proceedings. One idea is to introduce news over and above the stochastic evolution of the two packs. This would only work in a training environment rather than where money is used.
So, the idea would be for the News Person (call CN short for CNN) to control the flow of external information that will have an impact on the price of the final contract. The CN should therefore start rumours, for example "Research piece is coming out that suggests the market has run its course", or "I am hearing rumours that there is a break in supply". CN can start a rumour anytime and with anyone with the sole purpose of injecting some short-term spikes in the market.
Once CN has started a rumour, he can at any time announce the news on a whiteboard next to the pit and add a numbered card to the mix of cards except that it remain unturned. The direction of the move needs to match the bias of the rumour (e.g. supply shortage = positive move) size of the movement is at CN descretion and can be between -15 and +15 but should match the severity of the rumour.
It will be interesting to see how many instantly react to the rumour, also there is the outstanding question of whether the card should immediately be turned over. I am against it on the ground that this is a futures market.
Also I would have a senior person who has elected himself to maintain order as an idea candidate.
So, the idea would be for the News Person (call CN short for CNN) to control the flow of external information that will have an impact on the price of the final contract. The CN should therefore start rumours, for example "Research piece is coming out that suggests the market has run its course", or "I am hearing rumours that there is a break in supply". CN can start a rumour anytime and with anyone with the sole purpose of injecting some short-term spikes in the market.
Once CN has started a rumour, he can at any time announce the news on a whiteboard next to the pit and add a numbered card to the mix of cards except that it remain unturned. The direction of the move needs to match the bias of the rumour (e.g. supply shortage = positive move) size of the movement is at CN descretion and can be between -15 and +15 but should match the severity of the rumour.
It will be interesting to see how many instantly react to the rumour, also there is the outstanding question of whether the card should immediately be turned over. I am against it on the ground that this is a futures market.
Also I would have a senior person who has elected himself to maintain order as an idea candidate.
Friday, January 28, 2011
Stop Loss
If we are playing for money, the question arises on how you should cope with counteparty risk. Amongst friends this is clearly not an issue, but if we are trying to simulate trading conditions then we should introduce either margin calling or stop loss.
If we use a spreadsheet to track all settled trades, then it is easy to track PL during the course of the trading sessions. So one approach is for the person acting as the settlement agent does margin calls to all participants so that if they have run out of money, they will be forced to reduce their positions.
This approach seems the most optimal, but can be high maintenance if it involved movement of banknotes. Therefore if a judicator is available who can run the Clearhouse, then they can monitor PL and send out instructions to the exchange members (MM, IB and LT) to cut their positions by half during the following session.
HF can be included in this process for simplicity, however a more accurate event is that they are given, say £2000 in capital and once their PL has broken through this amount, the IBs will have to create an accord to unwind the trades a a level that leaves the HF with zero cash. This creation is what happens when an OTC counterparty defaults to multiple dealers. The MM should make a special trading session to facilitate unwinding of exposure.
Given that the LT are trading on their own account, they have limited capacity and should have a lesser amount of capital compared to the HF, say £1000. If a local is within 5 points of going bust, then the judicator must instruct the MM to take over the positions for a fee of 5 points and bankrupt the LT. At that point, the LT should try and make money by shining shoes, bringing coffee, or get captial funding from a HF.
If we use a spreadsheet to track all settled trades, then it is easy to track PL during the course of the trading sessions. So one approach is for the person acting as the settlement agent does margin calls to all participants so that if they have run out of money, they will be forced to reduce their positions.
This approach seems the most optimal, but can be high maintenance if it involved movement of banknotes. Therefore if a judicator is available who can run the Clearhouse, then they can monitor PL and send out instructions to the exchange members (MM, IB and LT) to cut their positions by half during the following session.
HF can be included in this process for simplicity, however a more accurate event is that they are given, say £2000 in capital and once their PL has broken through this amount, the IBs will have to create an accord to unwind the trades a a level that leaves the HF with zero cash. This creation is what happens when an OTC counterparty defaults to multiple dealers. The MM should make a special trading session to facilitate unwinding of exposure.
Given that the LT are trading on their own account, they have limited capacity and should have a lesser amount of capital compared to the HF, say £1000. If a local is within 5 points of going bust, then the judicator must instruct the MM to take over the positions for a fee of 5 points and bankrupt the LT. At that point, the LT should try and make money by shining shoes, bringing coffee, or get captial funding from a HF.
Subscribe to:
Posts (Atom)