# Online Trading as a Secretary Problem

## TL;DR

This paper studies an online trading problem where an intermediary matches buyers and sellers with adversarial values, providing algorithms that optimize welfare and gain-from-trade in a randomized setting.

## Contribution

It introduces competitive algorithms for online trading with adversarial values, focusing on welfare and gain-from-trade objectives.

## Key findings

- Developed algorithms with competitive guarantees for welfare.
- Designed algorithms with competitive guarantees for gain-from-trade.
- Addresses online trading with adversarial and random elements.

## Abstract

We consider the online problem in which an intermediary trades identical items with a sequence of n buyers and n sellers, each of unit demand. We assume that the values of the traders are selected by an adversary and the sequence is randomly permuted. We give competitive algorithms for two objectives: welfare and gain-from-trade.