Showing posts with label HFT. Show all posts
Showing posts with label HFT. Show all posts

Friday, August 28, 2015

28/8/15: Gold & Silver: Does technical analysis beat the market?


An interesting piece of research co-authored by Brian Lucey on efficacy of technical analysis in gold & silver markets: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2652637.

"This paper studies whether intraday technical trading rules produce significant payoffs in the gold and silver market using three popular moving average rules."

And the conclusions are (emphasis is mine): "The initial results show that the SMA, WMA and EMA trading rules generate significant negative payoffs using the parameters common in the literature in the high-frequency gold and silver markets. This suggests that there is no significant profit to be gained from technical trading in the gold and silver markets. However, our parameter sweep results show that there are a number of parameter combinations that generate significant profit in the gold market, but none in the silver market. Further, the best performing rules have different parameters to those used the existing literature. We show that longer run averages should be used by investors on intraday data and that investors need to employ different parameters when utilising technical analysis on daily and intraday data. In order to examine whether investors could have actually utilised the best performing rules, we perform an in- and out-of-sample test and show that only the SMA rule for gold generates significant profits in the in-sample as well as the out-of-sample period. All of the other best rules in the in-sample period generate either insignificant or negative payoffs in the out-of-sample period."

Monday, November 3, 2014

Wednesday, April 16, 2014

16/4/2014: EU Parliament Passes Bail-ins, SRM and MiFID2


So the EU Parliament voted in the three proposals relating to banking sector 'reforms' in the EU. These included

  1. SRM set up - a EUR55bn Single Resolution Fund to be used as the last line of defence in the future banking crises. Here is an earlier note on how effective that will be in stopping bank runs: http://trueeconomics.blogspot.ie/2013/12/11122013-europe-have-any-firepower-for.html You can see recent assessment of the directive by the IMF and myself here: http://trueeconomics.blogspot.ie/2013/03/1532013-imf-assessment-of-euro-area.html
  2. Bank Restructuring and Resolution Directive - a directive that amongst other things sets the first line of defence at shareholders' bail-ins, followed by debt and depositors' bail-ins. That's right, depositors' bail-ins. And do note, the rabbit hole doesn't stop there - see box out here: http://trueeconomics.blogspot.ie/2014/02/1822014-wither-irish-manufacturing-not.html
  3. Bail-ins are now not just 'on the horizon' but are de jure a law. All depositors over EUR100,000 (the maximum amount for which bank deposits guarantee is allowed) will be at risk. The law requires depositors and bondholders to absorb the second hit to the minimum of 8 percent of total bank's liabilities.
  4. As Reuters recently reported, non-performing loans not covered by provisions currently make up ca 1/3 of equity across the top 20 banks in the euro area. And that is after massive waves of deleveraging, recapitalisations and equity rebuilding that took place since 2008, or over the last 5 years plus. Now imagine the likelihood of the next crisis requiring exhaustion of equity to regulatory minimum and subsequent call on depositors. Pretty darn high. You can read Reuters analysis here: http://www.reuters.com/article/2014/04/15/us-banks-tests-provisions-idUSBREA3E07F20140415 and see this handy infographic: http://pdf.reuters.com/pdfnews/pdfnews.asp?i=43059c3bf0e37541&u=2014_04_14_11_05_b27f82d139834fd1a98af554e6aade90_PRIMARY.jpg
  5. In addition to the above, the Parliament also passed the directive establishing the European Securities and Markets Authority - a Paris-based body in charge of regulating trading and other activities in the markets, and the revamped amended MiFID2. This sets up the basic terms for regulating trading in securities and derivatives and contains new rules for trading commodities, OTC derivatives and HFT. ESMA is now empowered to write some 175 new rules to deploy MiFID2 and the proposals for these are due in May. One key area of regulation will cover HF traders, with all market traders requiring to register as either market participants or HF traders, as well as disclose their strategies.



Saturday, March 8, 2014

8/3/2014: FTT - More Benign Estimates of Impact?


In recent years, I have written extensively about the problems relating to the introduction of a Tobin-styled FTT, including as proposed by the european authorities.

Last year, I cooperated with an academic survey of the extent literature on FTT across various asset markets and instruments. Using meta analysis that study concluded that on the net, FTT will likely result in:
1) revenues well below those expected by the policymakers, and
2) significant reduction in markets efficiency and price discovery, including potential for adverse changes in liquidity risk environment in the markets for major financial instruments.

This February, a new working paper, titled "A General Financial Transactions Tax: Motives, Effects and Implementation According to the Proposal of the European Commission" by Stephan Schulmeister (WP: 461/2014 Österreichisches Institut für Wirtschaftsforschung, February 2014 Source: http://www.wifo.ac.at/wwa/pubid/47125) summed up "the main arguments in favour and against a FTT" and provided "empirical evidence about the movements of the most important asset prices."

The author shows that "long swings [in the asset prices] result from the accumulation of extremely short-term price runs over time. Therefore a (very) small FTT – between 0.1 and 0.01 percent – would mitigate price volatility not only over the short run but also over the long run."

In this, the paper conclusions are not novel.

It is generally accepted that efficiency-enhancing FTT will require extremely low rate of taxation in order to 'separate' HFT activities from long-only investment activities. The premise for this is well established in the literature: it is believed that higher order volatility in the markets is induced by HFTs and not by long-only or covered shorts positions.

Alas, I am not entirely convinced that we should be concerned with higher order volatility. Short-lived multiple-sigma events - capturing imagination of the media and the public - are not as disruptive as structural crises. And we all know that structural crises have nothing to do with either naked shorting, leveraged shorting or HFTs. These crises are not caused by the active trading. They are caused by active and sustained fraud or passive and sustained failure to enforce existent regulations, or both. On behavioural side, they are also caused by the 'exuberant expectations' - a situation where individuals mis-price directional risks. None of these causes is subject to FTT constraints if the tax is set at the levels where it is not impeding lucidity and price discovery.

So from the very top, the rationale presented in the paper to support FTT introduction (high frequency volatility) is distinct from the rationale presented by the EU leaders for introducing FTT (structural crises).

It is worth noting that Schulmeister puts heavy emphasis in the causality argument on the feed through from HFT to algos, relying on short shocks propagation mechanism via algos-induced changes in the trend.

The problem with this argument is that

  1. It ignores the existence of arbitrage opportunities (lack of contrarian algos is hardly consistent with Schulmeister's worldview)
  2. It also fails to account for reversion to the mean property of algos.


The paper "discusses the most important implementation issues if only a group of 11 EU member countries introduces this tax (without the UK). If London subsidiaries of banks established in one of the FTT countries are treated as part of their parent company, overall FTT revenues of the 11 FTT countries are estimated at € 65.8 billion, if London subsidiaries are treated as British financial institutions, tax revenues would amount to only € 28.3 billion."

The problem with the above that while the amounts are small, potential disruptions to the markets generated by, say, a 10bps tax can be significant. Take equities portfolio, returning 5% pa gross FTT will reduce the base by 0.1% or 0.2% on trade covered by a derivative contract. Thus, for full execute of a simple long-only strategy, involving simple one-direction hedge, the total tax exposure under the 0.1% FTT is 30 bps. Which is consistent with a 6 percent drop on gross return.

Thus, even if FTT were to deliver reduced short-term volatility, since long-only holders face a new tax, equivalent to roughly 1/5th of the CGT (if CGT is set at 30%). This is hardly immaterial.

Another issue arises in the context of the numerical estimates presented in the paper. The upper envelope estimate of EUR65.8 billion is based on the assumption of zero migration by institutions. EUR28.3 billion lower envelope estimate is based on the assumption that some migration is possible to the UK, with such migration triggering FTT application only to one side of trade (the side domiciled in FTT-imposing country). Alas, obviously, the exercise fully ignores the possibility to both sides of the trade migrating to non-FTT jurisdiction.

Saturday, October 19, 2013

19/10/2013: WLASze Part 2: Weekend Links on Arts, Sciences and zero economics


This is the second post of my WLASze: Weekend Links on Arts, Sciences, and zero economics for this week.

Enjoy and be warned, I do stray into 'some' economics (but only as 'science') in this one...


"By discovering a new set of solutions to the famous Maxwell equations governing electromagnetism, Hridesh Kedia of the University of Chicago and his colleagues have shown that light can be tied up in knots. Here the purple and gold cords represent the twisted magnetic field lines of knotted light." Ughh?.. No, really cool - read more here: http://www.scientificamerican.com/article.cfm?id=tying-light-in-knots-slide-show

And… "The same University of Chicago lab, led by William Irvine, also recently discovered a way to tie water up in knots. This photograph shows a basic knotted shape called a trefoil knot made of water, imaged by light scattering off tiny gas bubbles in the liquid."


Question for scientists… can anyone untie the knots of Irish policymakers' ideas? Like the following conjecture: to deal with the effects of the property bubble collapse, Budgets 2010-2014 introduced series of property markets tax incentives… Bet that'll be harder than Maxwell's equations…


But this is WLASze, so let us not dwell too long on matters of Irish policies. Even if for the purpose of advancing the science of the bizarre…

So back to sciences: "Society's techno-social systems are becoming ever faster and more computer-orientated. However, far from simply generating faster versions of existing behaviour, we show that this speed-up can generate a new behavioural regime as humans lose the ability to intervene in real time."

Ok, this is potent stuff. And where better to look for such 'machine outpaces mankind to defeat the entire purpose of the human-made system' than in financial markets (I can't shake off this year's 'Nobel' in Economics)… So: "Analyzing millisecond-scale data for the world's largest and most powerful techno-social system, the global financial market, we uncover an abrupt transition to a new all-machine phase characterized by large numbers of subsecond extreme events. The proliferation of these subsecond events shows an intriguing correlation with the onset of the system-wide financial collapse in 2008. Our findings are consistent with an emerging ecology of competitive machines featuring ‘crowds’ of predatory algorithms, and highlight the need for a new scientific theory of subsecond financial phenomena."

Here's the full article: http://www.nature.com/srep/2013/130911/srep02627/pdf/srep02627.pdf

Wanna see something really scary?


Or in different visualisation:



Ok, now that I am onto Financial Markets and their 'efficiency' - a compendium of links explaining this year's Nobel Memorial Prize in Economics. No commented from me:



Now I've done it… this was supposed to be WLASze and it is now more Finance & Economics than Arts or Sciences…

I recently wrote about the determinants of what makes content go viral on social networks (see link here). The study was based on Google+ - a network that Google seems to think is a major one, yet everyone else seems to think of as a 'may be one day' contender. Now, more real viral propagation visualisation via twitter:


Link: http://blogs.hbr.org/2013/08/visualizing-how-online-word-of/


As Apple is pushing ahead with the meat (USD5 billion) plan for new HQs in Cupertino, CA, here's an overview of the project: http://www.dezeen.com/2013/10/16/fosters-apple-campus-unanimously-approved-by-cupertino-city-council/


Yes, buildings of this scale are a challenge. Yes, aesthetics of the corporate identity at this scale are a challenge. Yes, Apple is strongly 'impositional' organisation with emotional attachment to 'eco-sthetics' and reality of a massive Death Star orb floating in alien space… so then, yes, the plans are perfectly befitting the client…

As a life-long fan of Apple (I still have a working-order first marketed laptop by the company in my collection and our household has a pile of interconnected Apple devices) all I can say is, sadly, with the shift in Apple's fortunes, the company is now running out of ideas… But wait, I am not alone: http://www.dezeen.com/2013/09/13/apple-has-reached-creative-saturation-says-steve-jobs-colleague-hartmut-esslinger/


On a beautiful side (to round off this week's WLASze): Gagosian Gallery is hosting a Willem De Kooning Ten paintings show, comes November 8:
http://www.gagosian.com/exhibitions/willem-de-kooning--november-08-2013

De Kooning is a master of light and contrast, depth and balancing extremal abstract expressionism with minimalism. His work is stark, striking, bright with simplified, distilled essence of countering colour and movement and space. These are his later works, right before his death in 1997.



The middle image above is taken from the coverage of the MoMA retrospective of de Kooning's works back in 2011.

It was not always thus, as you can see from his earlier works, such as Gotham News, 1955:


You can see more of de Kooning's works here: http://www.benditz.de/ and here: http://theartstack.com/artists/willem-de-kooning

Enjoy. De Kooning is a great master with deep, often disturbed depth of psychological insight alternating with organic capacity to surprise, to open up that momentary window into viewer's own imagination...

Monday, July 15, 2013

15/7/2013: Two notes on HFT effects on the markets


Two interesting notes on the financial markets general operational issues in relation to High Frequency Trading (HFT).


A quick post from the Aziz Economics: http://azizonomics.com/2013/06/15/have-financial-markets-gone-post-human/ on the topic of HFT and data disclosure. Do read the Nanex post cited: http://www.nanex.net/aqck2/4302.html

Basic idea is that speed of light separates trades in the current market. With some data being released in different formats and to different audiences at different times, this difference drives a massive wedge between HFT trades and ordinary order flows.

And a couple of quotes:

"... is having a two second jump on the market “insider trading”? Well, yes — but it’s legal insider trading with consent, out in the open."

Yep, you can pay more to get information ahead of everyone and then pay a bit more to execute trades ahead of the mere mortals. You can then collect the upside (you'd have to be pretty dim-witted to collect a downside on such a trade).

And that means that the old-fashioned elbow-grease and hard labour analysing stuff, forecasting it, setting a strategy, hedging etc… all become subservient to the speed of access + speed of execution.

Human is gone. Algo is in...
"… perhaps the beginning of the end for human traders is just the end of the beginning for global financial markets. Perhaps that is less of a death sentence, and more of a liberation, allowing talented human labour that in recent years has been channelled into unproductive and obscure projects in big finance to move into more productive domains."

I don't know. But I'd like to think a person is still somewhere under the sun in the markets. Otherwise, how can be make any connection between the financial markets, instruments traded and real side of the economy, aside from the sides glimpsed through high-frequency-advanced-release mechanism?..



And a paper on HFT effects on market index here: http://www.nature.com/srep/2013/130702/srep02110/full/srep02110.html

The paper shows that in short time scales stocks have a stronger influence on the index, rather than index has on stocks that are constituents of the index. This is encouraging, as it suggests that within shorter time horizons, extreme HFT-linked instrumentation of index is far less of a driver than HFT-linked instrumentation of individual stocks. In other words, whatever relevant information is contained in the stock gets indeed transmitted into index at high frequency and this information dominates index-own changes.

Monday, February 18, 2013

18/2/2013: Short-selling and Markets Volatility


A large number of analysts and policy makers tend to believe that highly leveraged trading activity, especially that linked to HFT, is a significant, even if only partial, driver of markets volatility. The channel through this logic usually works is that in the presence of leverage, speed of positions unwinding in response to unforeseen events increases, thus amplifying volatility.

An interesting study by Harrison Hong, Jeffrey D. Kubik and Tal Fishman, titled "Do arbitrageurs amplify economic shocks?" (Journal of Financial Economics, vol 103 number 3, March 2012, pages 454-470) examined the impact of arbitrageurs' activity on stock performance. Based on quarterly data from 1994 through 2007 for NYSE, Amex, and Nasdaq, share prices were examined over two distinct sub-periods: one day before earnings announcement and one day after the announcement. Medium-term performance was analysed for two days before earnings announcement and 126 days after earnings announcement.

The authors find that:

  1. Stock price reaction to earnings news is more severe in heavily shorted stocks than in stock with fewer short positions;
  2. Changes in the short ratio and earnings surprises counter-move;
  3. Share turnover as a result of large earnings surprises is higher for heavily shorted stocks as consistent with (1) above;
  4. Positive earnings surprises push up the valu of heavily shorted shares (as consistent with (1) and (2) above)
  5. Following positive earnings announcement, returns are higher (in general) for stocks with heavy shorting positions prior to the announcement since price appreciation post-announcement forces covering of short positions and triggers more demand for shares;
  6. Consistent with (5) above, post-positive earnings announcement, previously heavily shorted stocks become better targets for further shorting;
Overall, the study finds that:
  • Any earnings surprise in any direction (either positive or negative) leads to a corrective action by (either long or short) investors;
  • The above increases price sensitivity to newsflow and thus volatility;
  • Trading volume and stock price increase abnormally for heavily shorted stocks;
  • The abnormal volatility and volume & price effects are temporary and in the medium terms, prices revert to the mean.

Tuesday, October 23, 2012

23/10/2012: HFT restrictions and market efficiency


In my class on Investment Theory (MSc in Finance, TCD) we've discussed the issues relating to markets efficiency, HFT and relative speeds in newsflow and trading. We are going to talk more about this subject in my course on HFT in early 2013.

Here is the latest report on the effects of the EU regulatory interference in HFT.

Quote: "European Union plans to clamp down on trading shares faster than the blink of an eye could damage market efficiency and reduce liquidity, a UK government-sponsored paper said… A report by the Foresight Project, which was sponsored by the British government and gathered evidence from 150 academics and experts from 20 countries, said plans to force minimum resting times on orders could reduce liquidity."

The Project (led by John Beddington, the UK's chief scientific advisor) has found that:

  • "...some of the commonly held negative perceptions surrounding HFT are not supported by the available evidence and, indeed, that HFT may have modestly improved the functioning of markets in some respects"
  • "However it is believed that policymakers are justified in being concerned about the possible effects of HFT."
  • "The report found no direct evidence that HFT increased volatility, nor evidence to suggest it has led to an increase in market abuse."
  • "It said that computer-based trading could have adverse side effects in some circumstances and that these risks should be addressed."
As my students would know, I am of two views on HFT:
  1. HFT is a necessary activity with inherent risks (as any other activity in the market) 
  2. HFT can act in contradiction to the direct real-activity nature of the financial markets, but so can other financial instruments and strategies (e.g. hedging across non-asset-related risks, e.g. using Forex markets).