User:Alexkachanov/Finance/Technologies/Buy-side

каждой buy-side компании требуется свое ПО в зависимости от стратегии работы на рынке

When you trade actively (assuming you are filled, because someone may, of course, get their order in before you), you remove liquidity from the market, or take it out, so active trading is also sometimes known as take-out trading. HFT firms will employ specialized software programs for just such trading, and these programs are hence known as take-out engines or electronic eyes. Programs for passive trading are known by names such as quoting engines, autoquoters, and quote streamers.

Инвесторы

Portfolio management system
Execution management system - для исполнения ордеров при переделке портфолио - алгоритмическая торговая система, которая исполняет одрера на рынке, не двигая рынок. Конечно инвестор может воспользоваться и такой же системой, если ее предоставляет брокер, через которого инвестор работает с рынком.

в 2005 году еще даже различали trade management systems (TMSes) and order management systems(OMSes). В 2008 году - дебаты между Execution Management Systems (EMS) и order management systems (OMS)

In the early days of electronic trading, OMS and EMS enjoyed a state of peaceful co-existence. When traders sat at their computers, it was to trade and to do so they used an EMS. Portfolio managers and financial advisors reckoned which trades would best suit their clients by using the portfolio management and compliance tools their OMS provided.

The next step was to ask: Why not merge OMS and EMS and come up with something better? Not just bigger and better tools, but the killer app - a package that would integrate the best of both OMS and EMS.

The problems of integration, however, are embedded in the antagonistic natures of OMS and EMS and go back to their origins. The OMS model began in the back office; the EMS model with day traders. Generally, OMS is relatively costly and EMS relatively cheap. On the information technology level, OMS is heavy and EMS is light. OMS представляла собой клиент-сервеное приложение, устанавливаемое в компании, и поддерживаемое компанией. EMS работали по модели ASP. Какая-то сторонняя компания поддерживала сервер, а трейдеры пользовались клиентом у себя

wave of mergers and acquisitions aimed at integrating OMS and EMS platforms and ending the debate. In 2006, ITG purchased Macgregor in order to integrate its OMS, now dubbed Macgregor XIP, with the Macgregor Financial Network order-routing network. The resulting product has a market share of almost 30% in the space

The following year, 2007, saw a number of important deals. Linedata Systems, manufacturer of the LongView Trading OMS, said that its system would trade simultaneously with Goldman Sachs' REDIPlus. Fidessa, the offspring of buy-side OMS vendor LatentZero and sell-side EMS vendor Royalblue, offered its Minerva OEMS, or order execution management system

High frequency trading

Разновидность автоматизированной торговли. Не путать с алгоритмической торговлей.

Technologies

C++ and Java dominate, with a small number of FPGA specialists

Packet processing – TCP, UDP, Multicast
Destructor threads – spin waiting for packets to arrive to avoid interrupt wakeup delay
Actor model – minimise lock contention
Nanosecond timestamping
Direct buffer management
Warm up - allocate and preload everything before trading starts
Pinning memory, cache line alignment
CPU isolation and affinity
Achieving durability via replication across network rather than disk write (often to ‘luke warm’ backup server)
Memory mapping files when writing to disk cannot be avoided

C++ expertise area’s

Boost – particularly Math
C and even assembler inline
QuantLib - Greeks library
Lockfree++, actor patterns
Performance optimization with pragma’s
Intel TBB (Thread Building Blocks)
Code and data locality to reduce cache misses
Compiler optimization
Network processing – TCP, UDP, Multicast
TCP bypass - RDMA - libibverb
Memory optimization and tuning – cache alignment, huge pages
Tuning – Intel Vtune

Java expertise area

Low latency tuning and jitter avoidance
Extensive use of NIO, particularly with buffers
Lock detection, avoidance and tuning
Network processing - TCP, UDP, Multicast
Packet processing - raw, pcaps
RDMA – IBM JSOR, NIO wrappings for libibverb
GC tuning and avoidance - explicit buffer management and re-use, tuning -
- Numa aware on multi-socket servers -XX:+UseNuma
- Reduce the cache misses -XX:+UseCompressedOops
- Reduce the amount of TLB misses -XX:+UseLargePages
- Increase object persistence - -XX:MaxTenuringThreshold=4

Linux

Virtually all trading carried out on Linux

Goal is to achieve Low latency and low jitter
Programmers are expected to know about the techniques used since there are implications in the code
Constant battle with power saving features added to each new processor and Linux release - C and P states
Isolating cores from scheduler and then explicit core selection for critical threads - sched_setaffinity(2)
Turbo boost, overclocking
Kernel bypass preloads - Solarflare Onload, Mellanox libvma, speedus (in-memory transfer for applications running on same host and talking to each other using loopback)
TCP bypass - RDMA - RoCE, InfiniBand, OmniPath
Linux bypass - Data Plane Development Kit (DPDK)
Performance monitoring and profiling tools – sar, perf, iostats, mpstat, memprof, strace, ltrace, blktrace, valgrind, latencytop, systemtap ….

Strategies

Arbitration
- Exploiting difference in price for the same stock in different venues
Momentum
- Assumes that a current trend will continue
Alpha (pairs)
- Matches stocks and assumes the value of one should be the same as another
- Based on fundamental macroeconomic statistics
Composites
- Arb a ETF by beating it’s update, e.g. A 10% change in BP value on the FTSE100 may translate into a 0.2% change in the overall index
Market Making
- Accepting an Exchange fee to provide liquidity
- Typically large spread just outside current price
- Object is to make pennies on volume and minimize holding

Маркет-мейкеры

Раньше маркет-мейкинг делался вручную, теперь все больше этот процесс автоматизирован, так как ни одн человек не способен уследить за движениями рынка в наше время

In the options space, for example, reputable firms such as ORC, Actant, and RTS provide automated market-making software with all of the features you might need—quoting engines, electronic eyes, auto hedgers, and so on.

high-frequency trading system

Арбитражеры

high-frequency trading system

Предсказатели

Quantitative trading

cutting-edge advances in artificial intelligence to make its investment decisions.