## תקציר

Though the definition of gain from trade extends the definition of social welfare from auctions to markets, from a mathematical point of view the additional dimension added by gain from trade makes it much more difficult to design a gain from trade maximizing mechanism. This paper provides a means of understanding when a market designer can choose the easier path of maximizing social welfare rather than maximizing gain from trade. We provide and prove the first formula to convert a social welfare approximation bound to a gain from trade approximation bound that maintains the original order of approximation. This makes it possible to compare algorithms that approximate gain from trade with those that approximate social welfare. We evaluate the performance of our formula by using it to convert known social welfare approximation solutions to gain from trade approximation solutions. The performance of all known two-sided markets solutions (that implement truthfulness, IR, BB, and approximate efficiency) are benchmarked by both their theoretical approximation bound and their performance in practice. Surprisingly, we found that some social welfare solutions achieve a better gain from trade than other solutions designed to approximate gain from trade.

שפה מקורית | אנגלית |
---|---|

כותר פרסום המארח | Multi-Agent Systems - 16th European Conference, EUMAS 2018, Revised Selected Papers |

עורכים | Marija Slavkovik |

מוציא לאור | Springer Verlag |

עמודים | 109-126 |

מספר עמודים | 18 |

מסת"ב (מודפס) | 9783030141738 |

מזהי עצם דיגיטלי (DOIs) | |

סטטוס פרסום | פורסם - 2019 |

אירוע | 16th European Conference on Multi-Agent Systems, EUMAS 2018 - Bergen, נורווגיה משך הזמן: 6 דצמ׳ 2018 → 7 דצמ׳ 2018 |

### סדרות פרסומים

שם | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
---|---|

כרך | 11450 LNAI |

ISSN (מודפס) | 0302-9743 |

ISSN (אלקטרוני) | 1611-3349 |

### כנס

כנס | 16th European Conference on Multi-Agent Systems, EUMAS 2018 |
---|---|

מדינה/אזור | נורווגיה |

עיר | Bergen |

תקופה | 6/12/18 → 7/12/18 |

### הערה ביבליוגרפית

Publisher Copyright:© 2019, Springer Nature Switzerland AG.