-
Notifications
You must be signed in to change notification settings - Fork 26
/
Quantitative Finance.txt
2935 lines (2197 loc) · 132 KB
/
Quantitative Finance.txt
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
###########
### To Do
# develop flipped classroom model
I'm very interested in converting some of my materials to an online, interactive tutorial for the purpose of developing a flipped classroom model. Below is a link to my lecture materials.
The PDF files contain the narration text, formulas, and R code. Most slides have narration and formulas on the left, and the corresponding R code on the right.
For example you can take a look at time_series_multivariate.pdf
I also provide .R files containing the code, so that students can execute the code as they listen to the lecture. The lecture materials also contain .RData files with sample data.
https://drive.google.com/open?id=0Bxzva1I0t63vVGEtaXNIY1JMa00
# contribute to quant-lab
https://r-forge.r-project.org/projects/quant-lab/
# create website
https://wp.nyu.edu/
# get certification from AcademyR
http://www.revolutionanalytics.com/academyr-training-education
# Regression Models course
https://class.coursera.org/regmods-004/wiki/Lecture_note_links
https://github.com/algoquant/courses
http://www.rpubs.com/santam/logistic-regression
# write gitbook
https://www.gitbook.io/explore
# answer questions:
http://quant.stackexchange.com/questions/11413/how-to-compute-daily-skewness-of-sp-daily-return-timeseries-under-no-other-more
http://stats.stackexchange.com/questions/97783/understanding-vec2var-conversion-in-r
# needs an answer
http://quant.stackexchange.com/questions/9773/ib-with-r-which-package
###########
### quant top people resources
# Brian Peterson and Joshua Ulrich
Brian Peterson: Quantitative Strategy Development
https://r-forge.r-project.org/scm/viewvc.php/*checkout*/pkg/quantstrat/sandbox/backtest_musings/strat_dev_process.pdf?root=blotter
https://www.youtube.com/channel/UC0ZtUJOGGei4w5lxA_p8YKg
packages quantstrat, TTR, PerformanceAnalytics
Jeff Ryan packages xts, quantmod
Matt Dowle package data.table
# Aleksey Zemnitskiy Snowfall Systems PortfolioEffect
# top kaggle machine learning packages:
glmnet for Lasso and elastic-net regularized generalized linear models
gbm for gradient boosted ensemble models
e1071 for support vector machines C++ library libsvm
# Benjamin E. Van Vliet at IIT Chicago
http://stuart.iit.edu/faculty/benjamin-van-vliet
###########
### quant other people resources
# Attilio Meucci - KKR, SYMMYS, KEPOS
http://www.symmys.com
# The Thalesians
http://www.thalesians.com/finance/index.php/Main_Page
http://cran.stat.ucla.edu/web/views/Finance.html
http://matlab-trading.blogspot.ca
http://tradingwithpython.blogspot.com
http://www.statsblogs.com/tag/asset-allocation
http://quantivity.wordpress.com
# Empirical Finance
http://cran.r-project.org/web/views/Finance.html
# David Easley at Cornell: VPIN metric, volume clock, high frequency trading, machine learning
http://infosci.cornell.edu/faculty/david-easley
###########
### academics associations
# academic research websites top
http://blog.alphaarchitect.com/2015/04/01/where-to-find-cool-academic-finance-research/
# Q Group - Institute for Quantitative Research
http://www.q-group.org/
# Jack Treynor Prize Winners
http://www.q-group.org/jack-treynor-prize-winners/
# Real World Risk Institute by Nassim Taleb and Robert J. Frey (from Renaissance Technologies)
http://realworldrisk.com/
# Thalesians by Paul Bilokon at Deutsche Bank and Matthew Dixon in NYC
http://www.thalesians.com/finance/index.php/Main_Page
# Francis Diebold
http://fxdiebold.blogspot.com/
# Christopher Krauss Statistical Arbitrage Pairs Trading market anomalies momentum strategies value strategies
http://www.statistik.wiso.uni-erlangen.de/lehrstuhl/mitarbeiter/krauss.shtml
# Tim Leung University of Washington Pairs Trading
https://sites.google.com/site/timleungresearch/
# Thierry Roncalli - Kalman filter
http://www.thierry-roncalli.com/
# Andrea Frazzini - AQR & NYU
http://www.econ.yale.edu/~af227/
http://pages.stern.nyu.edu/~sternfin/afrazzini/index.htm
http://www.stern.nyu.edu/faculty/bio/andrea-frazzini
# Zura Kakushadze Quantigic Solutions
http://www.quantigic.com
# Nassim Nicholas Taleb
http://nassimtaleb.org/
http://www.fooledbyrandomness.com/
# Drew Creal Time-Varying Parameter Time Series GAS Models
http://faculty.chicagobooth.edu/drew.creal/
# Campbell R. Harvey backtesting data mining bonferroni adjustment
http://www.fuqua.duke.edu/faculty_research/faculty_directory/harvey/
# Wayne Ferson investment performance evaluation
http://www-bcf.usc.edu/~ferson/
# John H. Cochrane - Chicago Booth & AQR
http://faculty.chicagobooth.edu/john.cochrane/
# excellent courses, but no concise notes
http://faculty.chicagobooth.edu/john.cochrane/teaching/index.htm
# Notes for Time Series for Macroeconomics and Finance
http://faculty.chicagobooth.edu/john.cochrane/research/papers/time_series_book.pdf
# Asset Pricing
https://www.coursera.org/course/assetpricing
# Andrew Ang factors, anomalies
http://www.columbia.edu/~aa610/index.html
# Bruce E. Hansen - free econometrics manuscript
http://www.ssc.wisc.edu/~bhansen/
http://www.ssc.wisc.edu/~bhansen/crete/
# Ludwig Chincarini book Quantitative Equity Portfolio Management
http://www.usfca.edu/Faculty/Management/Ludwig_Chincarini/
# Matt Brigida teaching resources using R/Shiny
matt@complete-markets.com
http://complete-markets.com/sample-page/
https://github.com/Matt-Brigida
# Meese Rogoff Currency Forecasting Paper
http://www.ssc.wisc.edu/~bhansen/390/MeeseRogoff.pdf
# Robert Almgren - friend of Neil Chriss - Quantitative Brokers, high frequency volatility, order book
http://quantitativebrokers.com/
http://www.courant.nyu.edu/~almgren/
# Petter Kolm Director of Mathematics in Finance Program at NYU Courant
# teaches Algorithmic Trading and Quantitative Strategies
petter.kolm@nyu.edu
http://www.cims.nyu.edu/~kolm/
http://www.thestreet.com/story/10395354/1/how-nyu-plans-to-create-the-super-quant.html
# Peter Phillips (Yale professor) - unit root tests
http://korora.econ.yale.edu/phillips
# Alvaro Cartea - high frequency trading
http://www.cartea.net/
https://sites.google.com/site/alvarocartea/home/working_papers
http://scholar.google.com/citations?user=Kfl6fVQAAAAJ
http://www.ucl.ac.uk/maths/courses/msc-financial
https://sites.google.com/site/algorithmictradingbook/
# Sebastian Jaimungal - high frequency trading
http://www.utstat.utoronto.ca/sjaimung/
# Brandt Fuqua high frequency data intertemporal portfolio choice
http://people.duke.edu/~mbrandt/index.html
# Paolo Guasoni - Alpha Actively Managed Funds Optimal Consumption Utility very mathematical
http://www.guasoni.com/
http://www.bu.edu/math/people/faculty/mathematical-finance/guasoni/
http://www.dcu.ie/info/staff_member.php?id_no=3877
# Blake LeBaron agent based models nonlinear time series BDS test sofware
http://people.brandeis.edu/~blebaron/soft.html
# Santa Fe Artificial Stock Market simulation
http://artstkmkt.sourceforge.net/
The stock market simulation has agents following different trading rules.
The simulation exhibits bubbles and crashes.
The model finds an inverse relationship between diversity and asset price.
As an asset price reaches its peak, the diversity of trading rules
reaches its trough. (JP: I guess other trading rules are stop-lossed out?)
# Detecting Critical Transitions in Time series and Spatial data
http://www.early-warning-signals.org/home/
http://www.early-warning-signals.org/resources/code/
http://www.early-warning-signals.org/time-series-methods/metric-based-indicators/bds-test/
http://www.early-warning-signals.org/time-series-methods/metric-based-indicators/
http://www.early-warning-signals.org/time-series-methods/model-based-indicators/
# earlywarnings package
http://www.vasilisdakos.info/
http://www.early-warning-signals.org/resources/code/
https://github.com/earlywarningtoolbox
http://cran.r-project.org/web/packages/earlywarnings/index.html
# Wageningen University complex systems critical transitions
http://www.sparcs-center.org/
# Self-organized allocation to research crowdfunding democratization decentralization
http://www.sparcs-center.org/news/44/21/Self-organized-fund-allocation-SOFA.html
https://omslag.de/onderzoeksfinanciering/de-wijze-massa/
# William A. Brock economist Growth Theory Nonlinear Dynamics complex systems
https://en.wikipedia.org/wiki/William_A._Brock
# Riccardo Rebonato
http://www.marketswiki.com/mwiki/Riccardo_Rebonato
http://papers.ssrn.com/sol3/cf_dev/AbsByAuth.cfm?per_id=2146897
# German Creamer - Stevens Institute of Technology
http://web.stevens.edu/research/faculty_profile.php?faculty_id=1448
http://www.creamer-co.com/
# Carl Bacon book Practical Portfolio Performance Measurement and Attribution
# Andreas Steiner - consultant dislikes alpha
http://www.andreassteiner.net/
https://www.linkedin.com/in/andreassteiner
http://www.andreassteiner.net/consulting/
http://www.andreassteiner.net/consulting/publications.html
###########
### traders strategists investors
# Blair Hull CEO of Ketchum Trading
http://www.investopedia.com/news/3-trading-lessons-blair-hull/
# HTUS Hull Tactical ETF
https://finance.yahoo.com/quote/HTUS/?p=HTUS
# Jens Nordvig top currency and macro strategist at Bridgewater Associates founder of Exante Data
http://jensnordvig.com/
http://exantedata.com/
http://www.bloomberg.com/news/features/2016-09-15/wall-street-s-0-01-the-guru-who-only-talks-to-hedge-fund-elite
###########
### thoughts
I'm more interested in building an assembly line of quantitative trading models, rather than crafting a single trading model.
Bring quantitative trading models into the industrial age.
# Prerequisites for algorithmic trading strategies
Algorithmic trading requires a pipeline of strategies, not just a single strategy that looks promising.
Algorithmic trading requires a strategy lifecycle.
The critical question is when should a strategy be retired? If it hits a stop-loss? How to determine stop-loss amount? Using bootstrap of backtesting, to determine potential drawdowns and distribution of potential returns.
Strategies that trade infrequently can't really be tested properly.
Strategies that identify a few successful trades
High frequency data is needed for quant strategies for several reasons.
machine learning isn't effective for strategies that trade infrequently (weekly, monthly)
machine learning requires trading events
strategies which trade more often (intraday)
for learning
bias-variance tradeoff
Optimization in the presence of noise leads to spurious
# The reason skewness caught my attention as an indicator is because of the Amaya paper that I cited in my presentation. There's growing research showing that positive skewness in stock returns may predict their lower future returns. The analogue is buying options. When you buy an option you get positively skewed returns, but you must also pay an option premium, which lowers future returns.
# Finance theory states that it's not possible to obtain positive returns without taking risks (there's no free lunch)
One of the basic principles of finance theory is the relationship between risk and return on investment.
It doesn't mean that taking risk guarantees positive returns, because there are many strategies with lots of risk but no positive returns.
It means that the best performing strategies that do produce positive returns can't avoid taking risks.
It also means that strategies that don't take any risks can't produce positive returns.
# Capital Asset Pricing Model (CAPM) states that higher returns are only possible by taking more risk
We first select all the strategies that take the same level of risk, and then we select the best performing strategy from those strategies.
Then we compare the returns of these best performing strategies with their levels of risk.
The CAPM states that the returns of the best performing strategies are proportional to their levels of risk.
In other words, you can't obtain higher returns without taking more risk.
But the CAPM doesn't promise higher returns for taking more risk, since ther are many strategies with lots of risk but no positive returns.
# Finance theory states that arbitrage isn't possible in well functioning markets
Arbitrage means buying something at a lower price and immediately selling it at a higher price.
Arbitrage is a strategy that doesn't take any risks, and finance theory states that it therefore shouldn't be able to produce positive returns either.
Even though arbitrage shouldn't be possible in well functioning markets at equilibrium, arbitrage is often possible in real markets.
# Statistical arbitrage involves buying something at a lower price and selling it later at a higher price, or vice versa, first selling it and later buying it back at a lower price.
Statistical arbitrage is based on past statistical observations of price patterns, and occurs on a scale of several days or less.
Statistical arbitrage isn't riskless, since it involves owning a position for some time.
But statistical arbitrage is able to reduce risk on average over many trading cycles.
# Investors have a desire for assets with positive skewness, similar to lottery tickets (options).
This may be caused by
reduces their future returns.
low beta anomaly caused by demand for positive skewness (lottery) which reduces future returns
Bali Betting Against Beta Lottery Demand.pdf
Recent academic research is suggesting the existence of risk premia associated with higher moments of the return distribution.
For example skewness (third moment) is believed to predict lower future returns.
This may be caused by investor demand for assets with positive skewness (lottery tickets), which reduces their future returns.
We explore this phenomenon using high frequency data, since it allows for more accurate and timely estimation of higher order moments (bias-variance tradeoff).
Furthermore, we also explore the interaction between skewness and the momentum factor, since some recent research is suggesting that momentum returns are enhanced by skewness.
But using high frequency data can present a challenge, since the data size may cause R programs to run very slowly.
We demonstrate how to perform the analysis using vectorized R code designed for speed.
We present our results using interactive visualizations afforded by new R packages.
###########
### trading strategy ideas:
# is VIX trading volume able to forecast returns, or is volume just a proxy for volatility?
# VIX Futures term structure strategies
http://quantpedia.com/Screener/Details/198
# VIX/SPX trading strategies
http://www.bloomberglabs.com/quantitative-finance/research/vixspx-trading-strategies/
SPX mean reversion strategy using the spread between daily and weekly realized variance
VIX futures pair/spread strategy using the term-structure of VIX futures
# short SPX ATM puts strategy
http://www.thestreet.com/story/12770794/1/a-simple-options-trading-strategy-that-beats-the-sp-500.html
# test strategy where you exploit trending into market close:
jump on the trend in the last hour, then reverse position at the close, and reverse again at next day open
# perform portfolio optimization using two time scales (lookback periods)
estimate correlation matrix on a long lookback period
estimate returns on a short lookback period
the bias-variance tradeoff for correlation should be different than for returns
it's better to estimate returns on a short lookback period, but there's
no benefit to use short lookback period for estimating correlation
# perform running portfolio optimization assuming zero or constant correlations
update correlations over longer lookback periods
perform running portfolio optimizations over shorter lookback periods
invest in the portfolio based on its trailing return
use variance forecasts (instead of realized variance) to estimate portfolio variance
demonstrate that this portfolio is forecastable in-sample
demonstrate that it's also forecastable out-of-sample
explore shrinkage: does it improve forecastability?
explore long-only weights: that is just index (market portfolio)
explore mean reverting strategy: over intra-day period apply rule:
if realized return is higher than expected, then reduce long position, and vice versa
similar for expected negative return
# instead of optimizing portfolio for highest Sharpe, optimize for highest forecastability
is highest forecastability similar to lowest variance?
answer: highest Hurst exponent
can Hurst exponent be forecast?
create scatterplot of many portfolios: variance versus Hurst exponent
use package ForeCA by Georg M. Goerg at Google NYC
http://fastml.com/are-stocks-predictable/
http://stats.stackexchange.com/questions/126829/how-to-determine-forecastability-of-time-series
http://stats.stackexchange.com/questions/23007/assessing-forecastability-of-time-series?lq=1
calculate portfolio indicator similar to Sharpe: portfolio return divided by volatility
# forecastable component analysis package ForeCA
# find beta that maximizes Hurst
portfolio of stock plus beta times index
what is the beta that maximizes the portfolio's Hurst exponent?
is the beta stable over different intervals out-of-sample?
http://artax.karlin.mff.cuni.cz/r-help/library/ForeCA/html/ForeCA-package.html
# estimating Hurst exponent C++ code
http://www.bearcave.com/misl/misl_tech/wavelets/hurst/
http://www.bearcave.com/misl/misl_tech/wavelets/hurst/doc/index.html
# regress running Hurst exponent versus running volatility
regress running Hurst exponent versus returns over same lag
# if variance can be forecast, then forecasts can be used in pairs rebalancing trading
buy SPX and VIX and rebalance when SPX becomes cheap and vice versa
entry points are determined by forecasted volatility of SPX and VIX
# study periods of high volatility (variance) spikes
calcuate histogram of volatility - what's the distribution? chi-squared?
fit the chi-squared distribution using MoMoments
# identify periods around volatility spikes - say 100 bars before and after
what's the distribution of returns from period start to peak volatility,
and from peak volatility to period end?
# calculate Hurst exponent for volatility spikes, before min and after min
are returns autocorrelated around volatility spikes? yes
draw scatterplot of daily Hurst exponent versus volatility
draw scatterplot of Hurst exponent versus volatility spikes
create ARIMA model where autocorrelation term is weighted by level of VIX
create ARIMA model with correlation of returns to level of VIX
negative correlation means returns are skewed to downside when VIX is high
# calculate daily seasonality of Hurst exponent
# forecast stock index using VIX
high VIX produces high future returns
invest SPX notional inverse to VIX level
hysteresis: buy SPX at higher VIX, sell SPX at lower VIX,
Martin VIX Variance Stock Premium Forecasting.pdf
Bollerslev Stock Forecasting Volatility Market Premiums.pdf
Campbell Stock Forecasting.pdf
# create basic futures strategies:
# VIX/SPY
# VIX/VIX term
# VIX/TSY
# TSY/TSY term
Nagel Liquidity Premium Mean Reversion VIX Strategy.pdf
Donninger VIX Futures Basis Strategy.pdf
# VIX Futures ETF UVXY
http://www.bloomberg.com/news/articles/2016-04-22/record-vix-bets-keep-surging-as-wall-street-divines-mixed-signal
# study trading strategy that is short at period start, then flips long
and rides recovery
# forecast price bottom or volatility peak from moving averages
is there some scaling law?
# forecast variance using different models
variance can be forecast using GARCH
create indicator of variance trend from two VW Avg Variance
does the forecast variance also help to forecast returns?
# divide time into periods of low volatility and high volatility
calculate Hurst exponent in each period
# create trading strategy
thesis: volatility is forecastable,
returns are forecastable in periods of high volatility
identify when volatility will spike, and follow return trend in periods of high volatility
# perform GramSchmidt procedure to obtain PCA
repeat GramSchmidt procedure but maximize Hurst or SR, instead of variance
interpret the max Hurst components and observe how they change
regularize the procedure (shrinkage)
regress the max Hurst components against SPY
# draw scatterplot of daily SD versus daily returns
# create trading strategy with dual indicators
slow indicator for momentum and fast indicator for mean reversion
trade only on fast indicator, but bias its positions using slow indicator
when slow indicator indicates long position,
then fast indicator should switch between long and flat positions, and vice versa
# volatility reduction translates into growth
Kelly criterion
# volatility estimation
Stroud High Frequency Forecasting Volatility VIX VXX Strategy
# volatility pumping strategy
http://www-isl.stanford.edu/~cover/portfolio-theory.html
# Thomas Cover Universal Portfolios
https://en.wikipedia.org/wiki/Universal_portfolio_algorithm
Cover Universal Portfolios MathFin.pdf
Cover Universal Portfolios.pdf
Witte Trading Volatility Pumping Harvesting.pdf
Witte SaP500_HistPrices.R
Witte FTSE_DAX_DJIA_FollowWinners.R
Heaton Stock Selection Active Portfolio Management.pdf
Johannes Stock Forecasting Intertemporal Universal Portfolios Choice.pdf
http://optimallog.blogspot.com/2012/06/universal-portfolio-part-6.html
http://optimallog.blogspot.com/2012/07/universal-portfolio-part-9.html
http://optimallog.blogspot.com/2012/08/universal-portfolio-part-10.html
# Parrondo's paradox Brownian ratchet
https://en.wikipedia.org/wiki/Parrondo%27s_paradox
http://www.cut-the-knot.org/ctk/Parrondo.shtml
http://seneca.fis.ucm.es/parr/GAMES/Paradox%20in%20Game%20Theory%20Losing%20Strategy%20That%20Wins.htm
http://www.nytimes.com/2000/01/25/science/paradox-in-game-theory-losing-strategy-that-wins.html
http://parrondoparadox.blogspot.com.es/2011/02/parrondos-paradox-stock-market.html
http://quant.stackexchange.com/questions/1710/proof-that-you-cannot-beat-a-random-walk
http://quant.stackexchange.com/questions/352/volatility-pumping-in-practice
http://quant.stackexchange.com/questions/10887/harnessing-small-correlations-for-reliable-profit
# Brownian flashing ratchet
https://en.wikipedia.org/wiki/Brownian_ratchet
http://www.imsc.res.in/~sitabhra/research/persistence/page1.html
http://www.imsc.res.in/~sitabhra/research/persistence/page2.html
http://www.imsc.res.in/~sitabhra/research/persistence/page3.html
# risk parity strategy
First day of every month, sell all portfolio holdings at market. Same day, buy 50 random stocks from the S&P 500 constituents. Weight positions by inverse volatility, i.e. simple vola parity sizing. Always be fully invested.
http://www.followingthetrend.com/2015/06/a-random-ass-kicking-of-wall-street/
# risk parity by Bridgewater Associates
https://www.linkedin.com/pulse/our-thoughts-risk-parity-all-weather-ray-dalio?trk=hp-feed-article-title-ppl-follow
# use turning points in bearish percentage as buy/sell signals
http://www.zerohedge.com/news/2015-02-16/secular-death-market-bears-continues
# Two-Envelope Paradox
http://phys.org/news/2009-08-strategy-two-envelope-paradox.html
# send email to Josh at OneTick
# Random forest simple illustration
http://www.analyticsvidhya.com/blog/2014/06/introduction-random-forest-simplified/
http://www.analyticsvidhya.com/blog/2014/06/comparing-cart-random-forest-1/
# hosting Python and R models
https://yhathq.com/
https://github.com/yhat
# Raffael Vogler
http://www.joyofdata.de/blog/testing-linear-separability-linear-programming-r-glpk/
http://www.joyofdata.de/blog/category/machine-learning/
http://www.joyofdata.de/blog/
# Decision trees in Tableau using R
http://boraberan.wordpress.com/2014/02/07/decision-trees-in-tableau-using-r/
https://medium.com/code-poet/machine-learning-is-fun-80ea3ec3c471
# Replicate Paper with R and rCharts: AQR Dispelling Myths of Momentum
http://timelyportfolio.github.io/rCharts_factor_analytics/aqr_fact_fiction_momentum.html
https://github.com/timelyportfolio/rCharts_factor_analytics/
http://timelyportfolio.blogspot.com/2014/06/dispelling-myths-of-momentum-aqr.html
http://seekingalpha.com/article/2226863-myths-about-momentum-part-i
http://seekingalpha.com/article/2229023-myths-about-momentum-part-ii
http://quantstrattrader.wordpress.com/
# take Tour of Machine Learning Algorithms
http://machinelearningmastery.com/a-tour-of-machine-learning-algorithms/
# Build a Machine Learning Portfolio: Curate Your Models Like a Gallery
http://machinelearningmastery.com/build-a-machine-learning-portfolio/
http://machinelearningmastery.com/machine-learning-that-matters/
# explore publishing on:
https://medium.com/
###########
### Papers to read
Taleb Silent Risk.pdf
Cartea Algorithmic Trading with Learning.pdf
https://sites.google.com/site/algorithmictradingbook/
Qian Multiple Classifiers Stock Forecasting
German Creamer Logitboost Trading
Carr Optimal Trading Rules Without Backtesting
Rinne Seasonality Turn of the Month Stock Returns
DeMiguel Optimal Simple Portfolio Diversification
Prado Moment Skewness Kurtosis Gaussian Mixture Performance Tracking
Bailey Prado Strategy Overfitting Cross-validation
Bailey Prado Portfolio Optimization Algorithm
Faber Relative Strength Investment Strategies
Pav Markowitz Portfolio Distribution
Pav Markowitz Portfolio Signal Noise Ratio
Gatheral Random Matrix Theory Covariance Estimation
Bailey Prado Deflated Sharpe Ratio Overfitting
Daniel Momentum Crashes
Amaya Skewness Momentum Equity Returns
"We find a very strong negative relationship between realized skewness and next week's stock returns."
"A trading strategy that buys stocks in the lowest realized skewness decile and sells stocks in the highest realized skewness decile generates an average weekly return of 24 basis points with a t-statistic of 3:65."
Yang Volatility Filtering Technical Trading
Han Sorted Portfolios Trading Anomaly
Blitz Short-Term Residual Reversal
QUSMA ETF Daily Mean Reversion
Rudy Mean Reversion After Large Drops
Prado Backtesting Overfitting
Bailey Backtesting Overfitting
Bailey Backtesting Overfitting Cross-validation
Lo Adaptive Markets Hypothesis
Sutton Ensemble Learning Bagging Bootstrapping Boosting AdaBoost
Sutton Classification Evaluation Cross Validation Hypothesis Testing
Moskowitz Time Series Momentum
Bandarchuk Volatility and Momentum
Frazzini Buffetts Alpha
Asness Quality Minus Junk
Blitz Global Tactical Cross-Asset Allocation
Israel Size Value Momentum Anomalies
Keller Shrinkage Momentum Asset Allocation
Keller Momentum Markowitz Asset Allocation
Mitra Hurst Exponent Forecasting Financial Time Series
Grossmann Inverse Volatility Trading
Charpentier Bayesian Analysis Frequentist Hypothesis Testing.pdf
Masson Bayesian Analysis Frequentist Hypothesis Testing.pdf
###########
### suggest stackoverflow to dummies
You may want to submit your question to Stack Overflow:
http://stackoverflow.com/
In my opinion Stack Overflow is the best question and answer forum for programming.
But be sure to follow the rules of posting questions.
###########
### jobs in past
###########
### NYU Tandon School of Engineering
http://engineering.nyu.edu/academics/departments/finance/people
http://www.nyu.edu/search.directory.html?search=name%3Dpawlowski%26location%3DNew+York
http://engineering.nyu.edu/life/student-resources/registrar/academic-calendar#fall14
# Roy S. Freedman Inductive Solutions Artificial Intelligence Compliance Credit Interest Rate Models
http://en.wikipedia.org/wiki/Roy_S._Freedman
http://www.inductive.com/index.htm
http://www.inductive.com/rsfbio.htm
###########
### people I know
# Jonathan Stein
http://www.energyriskusa.com/static/jonathan-stein
###########
### R/Finance Chicago
### R/Finance Chicago 2016 best presentations
Jason Foster
Douglas Service
Qiang Kou
Piotr Orlowski
Bryan Lewis
Mark Bennett
Matt Brigida
Michael Kane
Xiao Qiao
Frank Diebold
Eran Raviv
###########
### news articles
# news blog data c++
https://tfetimes.com/
# (09-12-16) Volatility targeting and risk parity quant funds may be forced to sell in a market selloff as volatility andd correlations increase
http://www.bloomberg.com/news/articles/2016-09-12/a-sell-off-pressure-larger-than-brexit-imperils-one-of-wall-street-s-hottest-trades
# (09-09-16) Bridgewater Pure Alpha fund has declined -9.4% in 2016, while All Weather gained +13.5%
http://www.bloomberg.com/news/articles/2016-09-11/bridgewater-attracts-22-5-billion-financial-times-says
The Pure Alpha fund has returned about 12% per year since 1991, and hasnt had a losing year in the last 15.
The Pure Alpha fund owns $69 billion of assets.
Despite the loss, the Pure Alpha fund is receiving $22.5 billion from investors, after it opened to new money.
# (09-09-16) Investors in 2016 have redeemed the most capital from traditional hedge funds since 2008
http://www.bloomberg.com/news/articles/2016-09-09/perry-capital-assets-plunge-60-to-4-billion-as-wagers-backfire
# (09-07-16) Banks and investment funds are hiring quants
http://www.bloomberg.com/news/articles/2016-09-09/ubs-wealth-management-s-haefele-joins-battle-to-recruit-quants
# (09-01-16) Jaffray Woodriff and David Vogel are using machine learning and their Voloridge and QIM funds are strongly outperforming
http://www.bloomberg.com/news/articles/2016-09-01/code-bros-united-by-netflix-run-two-of-world-s-best-quant-funds
# (08-21-16) Simplex Equity Futures aglorithmic fund in Tokyo is outperforming other hedge funds
http://www.bloomberg.com/news/articles/2016-08-21/hedge-fund-robot-outsmarts-human-master-as-ai-passes-brexit-test
The Simplex fund is up 1.9% from its inception in April 2016 through August 19th, while the Topix index has lost 16% in 2016.
# (08-17-16) Quantitative funds are attracting capital as they continue to outperform other hedge funds in 2016
http://www.bloomberg.com/news/articles/2016-08-21/hedge-fund-robot-outsmarts-human-master-as-ai-passes-brexit-test
Two Sigma was founded by David Siegel and John Overdeck, and now manages $35 billion.
David Siegel worked for Tudor Investment, and Two Sigma was seeded by Paul Tudor Jones in 2001.
Steven Cohen invested $250 million in Quantopian.
# (07-15-16) S&P500 has reached a new high but defensive stocks are leading the rally, and stock picking funds have not profited despite low stock correlation
http://www.ft.com/cms/s/0/ede9f018-49cf-11e6-b387-64ab0a67014c.html
Both stocks and bonds have rallied in anticipation of further low interest rates and central bank stimulus.
Low volatility stocks and high-yielding stocks have outperformed, because they are substitutes for bonds.
Defensive stocks such as utilities, real estate and telecoms have also outperformed, which is typical for the end of the business cycle, before profits fall and a recession starts.
Stock correlation has dropped since May 2015, with 109 stocks gaining at least 20%, and 96 falling at least 20%.
But stock picking funds have not profited, with equity hedge funds producing a small negative return in 2015, and a small negative return so far in 2016.
# (07-01-16) All assets (stocks, bonds, and commodities) rallied after Brexit, because markets are expecting central banks to be dovish
http://www.bloomberg.com/news/articles/2016-07-01/everyone-s-a-winner-in-brexit-aftermath-as-doves-rescue-market
# (06-30-16) Soros says Europe is in a deflationary trap that will get worse because of Brexit
http://www.bloomberg.com/news/articles/2016-06-30/soros-says-brexit-has-unleashed-a-financial-markets-crisis
The EU will disintegrate unless voters see that the EU is benefiting their lives.
# (06-29-16) British stocks have erased all Brexit losses, and have outperformed eurozone stocks
http://www.zerohedge.com/news/2016-06-29/brexiteers-1-0-scaremongers-uk-stocks-erase-all-brexit-losses
http://www.bloomberg.com/news/articles/2016-06-29/it-took-just-four-days-for-british-megacaps-to-shake-off-brexit
If British stocks continue outperforming, then it will create a big incentive for other EU members to exit as well.
# (06-24-16) Soros didn't short British pound before Brexit vote, but he shorted stocks and bought gold, which both made him money
http://www.marketwatch.com/story/soros-looks-set-to-make-a-killing-on-brexit-result-2016-06-24
http://www.bloomberg.com/news/articles/2016-06-27/soros-was-long-the-pound-before-brexit-vote-says-spokesman
# (06-24-16) gold rallied up to 8% after Brexit vote
http://www.bloomberg.com/news/articles/2016-06-26/hedge-funds-win-world-beating-rally-with-record-gold-holdings
# Brexit is extremely negative for European markets
http://www.zerohedge.com/news/2016-06-24/who-are-biggest-losers-brexit
The British stock market was the best performing European stock market on Friday after the Brexit vote.
# (06-15-16) Quantitative funds are outperforming other hedge funds in 1H16, up about 8% versus other hedge funds flat
http://www.bloomberg.com/news/articles/2016-06-15/machines-beat-humans-in-hedge-fund-quest-to-time-market-bottom
The biggest shift in momentum was in energy stocks.
Quantitative funds identified long energy stock trades both based on valuation and on momentum.
# (05-30-16) Quantitative equity funds have had strong returns, and now use big data techniques
http://www.pionline.com/article/20160530/PRINT/305309990/quants-throw-tradition-out-the-window
Acadian Asset Management $9.2 billion global managed volatility equity strategy beat its MSCI World index benchmark by 12.63% for the 12 months through March 31, while Analytic's $3.8 billion global low-volatility equity strategy exceeded its benchmark by 9.01% and Martingale's U.S. low-volatility large-cap equity strategy added 8.2% for the same period.
# (05-31-16) Value stocks have underperformed growth stocks since 2006, the longest losing stretch in history
http://www.bloomberg.com/news/articles/2016-05-31/lost-decade-for-value-stocks-tests-faithful-who-say-end-is-nigh
In the 1990s growth stocks outperformed drastically, then in 2000-2006 value stocks outperformed drastically.
Currently banks make up about 30% of large-cap value stocks, and their underperformance since 2006 largely explains the underperformance of value stocks.
Low interest rates have held back bank stocks, and the repeated failure of rate hikes have kept bank stock prices low.
But 2016 may be a turning point for value stocks.
Value stocks in the Russell 1000 Index have gained 4.3% in 2016, compared with 1.1% for growth and 2.7% in the S&P500.
The KBW Bank Index has jumped 18% in the past three months, double the increase of the S&P500.
Strong global economic growth is needed for value stocks to outperform.
# (05-25-16) Quant fund XTX Markets is fourth biggest foreign exchange trader, with 7.6% of spot volume, ahead of Deutsche Bank
http://www.bloomberg.com/news/articles/2016-05-25/computerized-trader-leapfrogs-deutsche-bank-in-currency-rankings
XTX also makes markets in commodities, stocks and derivatives.
# (05-15-16) Quant hedge funds have been outperforming macro, merger arbitrage, and other funds in 2015 and 2016
https://www.theguardian.com/business/us-money-blog/2016/may/15/hedge-fund-managers-algorithms-robots-investment-tips
Stock picking by human experts has become very risky.
David Einhorn recommended buying SunEdison before it filed for bankruptcy.
Bill Ackman promote Valeant before its stock crashed.
David Siegel, cofounder of Two Sigma Investments, stated that one day no human investment manager will be able to beat the computer algorithm.
# (05-12-16) Clifford Asness: hedge fund returns are very correlated to the market, and therefore don't jsutify their high fees
http://www.bloomberg.com/view/articles/2016-05-12/hedging-on-the-case-against-hedge-funds
Hedge funds have a beta of about 35% to 40%.
Hedge funds need to hedge more of their beta, and charge lower fees.
# Hedge funds lost 1.9% in 1Q2016, the worst start since 2008
http://www.bloomberg.com/news/articles/2016-05-11/rubenstein-says-he-s-surprised-macro-hedge-funds-got-it-wrong
979 funds closed in 2015, more than any year since 2009.
# Manoj Narang creates new firm to combine high frequency trading with statistical arbitrage strategies
http://www.bloomberg.com/news/articles/2016-03-11/a-high-speed-trading-pioneer-spots-a-new-way-to-make-big-returns
High frequency trading relies on "structural" information about the order book, and profits mostly on front-running other traders by guessing their intentions.
High frequency trading uses prices and order book data at around milisecond frequencies, and holds positions for seconds or less, and requires very fast connections to exchanges.
Statistical arbitrage identifies statistical anomalies in asset prices, and profits by betting that those anomalies will revert back to equilibrium values.
Traditional statistical arbitrage uses daily prices to identify price anomalies, and holds positions for at least several days.
Manoj Narang is a veteran high frequency trader who recently left Tradeworx.
Manoj Narang's new fund will focus on volatility arbitrage.
# (04-05-16) Stock picking funds in 1Q16 have underperformed the market by the most in two decades
http://www.bloomberg.com/news/articles/2016-04-05/lockstep-moves-ease-futility-doesn-t-for-u-s-equity-managers
Correlation between stocks is at the lowest levels since 2012, which should be good for stock pickers.
But stock picking funds owned the wrong stocks, because the stocks that were unloved by stock picking funds outperformed the ones they loved by the most since 2013, with lower quality stocks rallying.
Portfolio managers made a mistake by avoiding utility and consumer staples stocks, while buying financial and technology stocks.
Stocks that performed the worst in 2015 rallied by 8.2% in the first quarter of 2016.
Momentum strategies also lost money because trends quickly reversed, as the S&P first lost 10% but then recovered.
# (04-06-16) Tiger Global hedge fund plunged -22% in the first quarter 2016.
http://www.wsj.com/articles/tiger-global-hedge-fund-plunges-on-amazon-netflix-other-bets-1459961411
# (03-28-14) Kyle Bass fed negative rumors to CNBC while selling short mortgages
http://www.wsj.com/articles/goldman-sachs-and-bear-stearns-a-financial-crisis-mystery-is-solved-1459205026
On Wednesday, March 12, 2008, CNBC reporter David Faber asked Bear Stearns chief Alan Schwartz if it was true that Goldman Sachs wouldnt accept Bear Stearns counterparty risk.
David Faber got the story from Kyle Bass.
This rumor led to the collapse of Bear Stearns.
# (03-03-14) Asness: Great Divide over Market Efficiency
http://www.institutionalinvestor.com/Article/3315202/Asset-Management-Equities/The-Great-Divide-over-Market-Efficiency.html
# (03-24-16) Most active bond traders are losing money in 2015-16, despite bonds rallying
http://www.bloomberg.com/gadfly/articles/2016-03-24/if-every-bond-s-a-winner-why-are-there-so-many-losers
The Merrill Lynch Global Bond Index has gained almost 3% in 2016, the biggest gain in the same period since 1997.
But the returns at nontraditional bond funds are slightly negative on average in 2016, and some have lost a lot.
The Altegris Fixed Income fund is down more than 8% and the Highland Credit Fund has lost more than 6% year to date 2016.
European hedge funds have struggled, with Oceanwood Capital $2.1 billion hedge fund losing 8.1% through March 12, 2016.
The winners are central banks who own the sovereign bonds that have experienced the biggest gains.
# Small quant funds grow by using open-source software and cheap data
http://www.bloomberg.com/news/articles/2016-03-16/barbarian-coders-at-the-gate-the-inexorable-rise-of-diy-quants
Huge leaps in computing power are transforming modern finance in ways that few have ever imagined and giving tech-savvy, DIY types the tools to compete with large quantitative funds.
Emanuel Derman says there's so much available data, open-source software, and computing power, that samll quant funds can get started in no time at all.
Open-source software written in R and Python is lowering the barriers to entry for small quant funds.
But growth of quant funds is leading to crowded trades which may have caused the August 24th 2015 flash crash.
Ernie Chan and Roger Hunter run QTS Capital Management.
QTS buys futures tick data from AlgoSeek for $500 a month.
QTS returned 12% in 2015, outperforming the U.S. stock market and the average hedge funds.
Since 2014 QTS assets have grown more than fivefold to $22 million.
On any given day QTS is testing 10 different models while executing eight trading strategies.
Quantopian started in 2011 and its membership has increased from 1,570 to more than 66,000 in 2016.
Estimize provides crowd-source corporate earnings estimates.
# Peter Muller hedge fund PDT
http://www.forbes.com/sites/nathanvardi/2016/01/04/the-new-quant-hedge-fund-master/
# David Siegel and John Overdeck hedge fund Two Sigma
http://www.forbes.com/sites/nathanvardi/2015/09/29/rich-formula-math-and-computer-wizards-now-billionaires-thanks-to-quant-trading-secrets/
John Overdeck worked for D.E. Shaw and Amazon.
Two Sigmas two main funds are Eclipse and Spectrum. Eclipse has a shorter holding period of several weeks, while Spectrum has a holding period closer to one month.
Two Sigma is well known to pursue in court its employees for stealing proprietary models or data.
Two Sigma accused researcher Jianjun Qiu of stealing their proprietary models or data. Qiu escaped to China, reached a settlement with Two Sigma, and now works for Citadel.
# The flash rally in U.S. Treasuries on October 15, 2014 was caused by illiquidity in cash market
http://www.bloomberg.com/news/articles/2016-02-26/jpmorgan-s-flash-rally-theory-contains-message-on-today-s-market
# (02-28-16) Early 2016 selloff led by most popular analyst stock picks
http://www.bloomberg.com/news/articles/2016-02-28/analyst-darlings-turn-stock-market-pariahs-as-momentum-unravels
Stocks ranked highest by analysts have fallen on average 11% in 2016, while stocks ranked lowest are down only 3.4%.
This is tied to the reversal of high momentum stocks, many of which are owned by hedge funds.
The stocks that are the most owned by hedge funds are trailing the S&P500 by 4.6% in 2016, after outperforming by an average of 10% per year over the last four years.
In March 2009 analysts were bearish when it was the ideal time to buy stocks.
# (02-23-16) Hedge funds underperformance in early 2016 caused by large exposure to momentum factor, which has reversed
http://www.zerohedge.com/news/2016-02-23/how-outperform-most-hedge-funds-2016
Hedge funds have large positions in high momentum stocks, they have large exposure to the momentum factor.
Goldman Sachs created a Hedge Fund VIP basket of stocks to which many hedge funds large exposures (Bloomberg ticker: GSTHHVIP).
Many hedge funds own similar stocks, and most of those belong to the VIP basket.
Many hedge funds are also very concentrated in a small number of stocks, with the average fund holding 68% of long assets in its top 10 positions.
# Trading firms and hedge funds are hiring quants who combine skills in trading, modeling and programming
http://www.nytimes.com/2016/02/23/business/dealbook/a-new-breed-of-trader-on-wall-street-coders-with-a-phd.html
https://www.janestreet.com/
Quant firms have prospered and grown, creating demand for traders who can build financial models and write computer code, and who can also spot market anomalies and bet big on them.
Issuing new ETFs has become a major profit center for large asset management firms like BlackRock, Vanguard and Invesco.
Jane Street specializes in trading ETFs, and has increased its shareholder equity from $228 million in 2007 to more than $1 billion in 2015.
Jane Street employs only 450 people in offices in New York, London and Hong Kong.
Yaron Minsky is the CTO of Jane Street.
Sandor Lehoczky is a star trader at Jane Street and chief quant.
# (02-22-16) Arnott Research Affiliates: value stocks are cheapest ever, besides the Nifty Fifty craze of 1970s
http://www.bloomberg.com/gadfly/articles/2016-02-22/don-t-be-tricked-by-the-alpha-mirage-from-smart-beta
# (01-27-16) Hedge fund underperforming massively, but investors still want them
http://www.bloombergview.com/articles/2016-01-27/investors-search-for-new-love-in-hedge-funds
http://www.bloomberg.com/news/articles/2016-01-26/bridgewater-s-dalio-trumps-soros-as-most-profitable-hedge-fund
The 20 most profitable hedge funds earned $15 billion in 2015, while the rest collectively lost $99 billion.
The top funds have earned 48% of the $835 billion in profits that the hedge fund industry has generated since its inception.
# (01-20-16) Best and worst stocks are diverging the most since the 1990s
http://www.bloomberg.com/news/articles/2016-01-21/short-sellers-are-making-an-even-bigger-killing-than-you-think
The top 10% most costly to short underperformed the market by a record 26% in 2015, which is the widest split between winners and losers since the 1990s. Even as the S&P dropped only 0.7% in 2015, the average decline in the 10 worst stocks was 64%, while the 10 best surged 71%. This divergence is mostly due to commodity companies underperforming and some technology companies outperforming.
# Many quant indexes have underperformed compared to their backtests
Beware of optimistic results of backtesting hypothetical investment strategies! Past backtesting results are no guarantee of future performance, as they say.
http://www.bloomberg.com/news/articles/2016-01-21/how-wall-street-finds-new-ways-to-sell-old-opaque-products-to-retail-investors
# Market depth is dropping despite high trade volumes
http://www.bloomberg.com/gadfly/articles/2016-01-15/robots-are-eating-your-retirement-in-volatile-stock-market
Market depth is how many contracts can be traded before moving the market by 1 point.
Market depth has declined by more than 60% over the past 2 years, as was illustrated during the August 24th 2015 crash.
Most trading is done by algos, even the trades from actively managed funds.
# (01-14-16) AQR ETFs are performing very well
http://www.bloomberg.com/news/articles/2016-01-14/cliff-asness-s-alternative-funds-top-rivals-with-17-returns
QMNIX AQR Equity Market Neutral Fund
QLEIX AQR Long-Short Equity Fund
# artificial intelligence machine learning startups in finance
http://themarketmogul.com/ai-and-finance-no-room-for-philosophy/
Kensho raised USD 15 million in order to train computers to replace financial analysts.
# Marko Kolanovic head JPM Quants making well timed market calls
http://www.zerohedge.com/news/2015-11-13/jpmorgans-gandalf-quant-nailed-it-again
http://www.bloomberg.com/news/articles/2015-09-25/it-s-marko-kolanovic-s-market-we-just-trade-in-it
# Most active managers underperform because they risk missing a few outperforming stocks
http://www.bloombergview.com/articles/2015-11-11/why-indexing-beats-stock-picking
Stock index returns are driven by a small group of the best performing stocks
in the index, while most of the rest perform poorly.
Active managers risk missing those outperforming stocks.
Active manager underperformance can be explained by creating random portfolios.
Most random portfolios underperform the index because they're missing the
outperforming stocks.
# (08-18-15) momentum strategies losing steam in Summer 2015
http://www.bloomberg.com/news/articles/2015-08-18/momentum-trade-gives-evidence-of-its-own-mortality-a-down-week
Since July 2015 industry leadership in the S&P 500 has been shifting to utilities, which are leading the market since August, and 2014 winners, like health-care and consumer companies, are trailing.
The Newedge CTA Index has fallen 6.5% from a high in April 2015.
Companies with strong balance sheets are rallying.
An index of stocks with the strongest balance sheets reached its highest level versus stocks with weaker ones since October 2013.
# (07-20-15) momentum strategies gained 15% in 2015
http://www.bloomberg.com/news/articles/2015-07-20/momentum-rules-in-u-s-stock-market-stuck-in-worst-rut-on-record
Momentum strategy of buying stocks with best price appreciation in recent quarters,
and shorting the worst have gained 15% in 2015.
The top three industries in the S&P500 in 2015 are retailers, health-care,
and biotech, which also led the market in 4Q2014, rising an average 14%.
While the S&P500 has not been up or down more than 3.5%, the difference in
return between the best and worst-performing industries is widening,
reaching the highest level since the bull market began in 2009.
Momentum strategies have beat every other quantitative strategies,
including growth, value, and size, which has only happened two other times in
24 years of data, in 2007 and 1993.
Momentum strategies work well in a maturing bull market such as now.
# algorithmic trading for retail investors is becoming popular
http://www.wsj.com/articles/an-algo-and-a-dream-for-day-traders-1439160100
# CFM crash predictor equals volatility divided by square root of trading volume
http://www.bloombergview.com/articles/2015-04-06/bitcoin-and-market-crashes
# Capital Fund Management (CFM) Stratus fund returned 17.7% through October 2014, second-best large hedge fund in Europe
http://www.bloomberg.com/news/2015-01-06/the-quants-on-the-left-bank-how-paris-based-physicists-used-algorithms-to-get-big-returns.html
The return of volatility in 2014 has been good for quants,
Seven quantitative funds are in the top 25 funds in 2014,
Man Group stock rose 74.6% in 2014,
# Jean-Philippe Bouchaud - Capital Fund Management
https://www.cfm.fr/
# hedge funds ranking in 2014: Many quant funds showed good performance in 2014: AHL, Two-Sigma, Citadel, Graham, D.E. Shaw, etc.
http://www.bloomberg.com/news/2015-01-04/credit-hedge-funds-game-stupid-wall-street-to-top-2014-ranking.html
# more hedge fund losses
# They couldn't make money because volatility was too low.
# Now they can't make money because volatility is too high.
# "Volatility is too damn high!"
http://www.bloomberg.com/news/2014-11-09/macro-funds-that-lamented-boring-market-lose-in-october.html
http://www.businessweek.com/articles/2013-06-06/how-the-robots-lost-high-frequency-tradings-rise-and-fall
http://quant.stackexchange.com/questions/1628/what-are-the-best-sources-for-equity-quantitative-research
http://quant.stackexchange.com/questions/1985/what-papers-have-progressed-the-field-of-quantitative-finance-in-recent-years-p
# there may be bursts of dependence leading to large cumulative losses
http://www.hussmanfunds.com/wmc/wmc141020.htm
###########
### behavioral finance
# Bloomberg Tradebook Trader Exercises
https://traderbrainexercise.com/
Denise Shull, founder of ReThink Group, New York consulting firm that coaches financial professionals and athletes:
Superior judgment of risk requires multiple cognitive and emotional abilities, that allow acquiring knowledge, recognition of reality, and understanding.
These include the ability to create mental models to predict prices, emotional self-awareness (knowing when emotions interfere with good judgment), and the ability to read other people's minds.
# Andrew Lo experiments show that good traders have more control over their emotions, and are flexible in response to risk and market volatility
http://www.bloomberg.com/news/articles/2016-09-01/wall-street-s-next-frontier-is-hacking-into-emotions-of-traders
Andrew Lo observed volunteers in a simulated trading game, and measured the trader's stress levels by measuring their perspiration and heart rate.
Lo's study found that good traders responded to stressful situations quickly, but afterwards calmed down quickly as well, leaving them better prepared for future challenges.
On the other hand, poor traders remained emotionally charged even after the volatility subsided.
Los study suggests that theres a sweet spot for emotional engagement: too much, and youre overly aggressive or fearful; too little, and you arent involved enough to care.
Having too much or too little emotion made for poor trades.
# Caltech researcher Peter Bossaerts found that traders use "theory of mind" when they make trading decisions
http://www.bloomberg.com/news/articles/2016-08-19/neuroscience-may-help-train-your-brain-for-trading
fMRI scans show that parts of the brain associated with the "theory of mind" are activated when traders try to forecast prices.
Traders that were better at predicting prices were also better on other tests of "theory of mind" abilities.
# People who are unsure and slow to make judgments are better decision makers
https://hbr.org/2016/07/slow-deciders-make-better-strategists
Researchers created a competitive strategy game, and analyzed the performance of different types of participants, based on their speed and confidence.