Here is a template algorithm from the insiders challenge you can use.

Model 0.1

Model #1

Model #2

Model #3

Model #4

Model #5

Model #6

Model #7

@Thomas,

If we want to include the Insiders dataset (two year holdout period), can we use the period January 4, 2014 - January 4, 2018 instead?

Very excited to have a go at this challenge!

@Joakim: Good question, yes absolutely. It won't matter for the scoring as we can run it internally on the full IS and OOS period, so the only disadvantage is to you in not being able to see part of the backtest period other users have access to.

model v0.1

Thomas, I posted this concern on the tearsheet thread -- this tearsheet doesn't factor in beta as a risk factor? So if I submit a short-only algorithm, for instance, the specific IR will be much lower than it should be.

model v0.2

Thanks @Anotony, I got mixed up on my notebooks!

factset estimates

alternative (self-serve) SHORT-ONLY

as mentioned earlier, the correct specific IR is closer to 2.0 than 1.0.

Factor was created > 6 months ago.

Model #8

Model #9

Model #10

Model #11

Model 1A

Model 1B (alternative ranking).

Model 2A.

Model 1

model 1

Model #1

Submission v1. Fixed some coding issues with my previous v1 submission to better reflect my economic hypothesis.

Thank you for all the submissions so far! One pattern I see here in submissions by Mikhai, Arun, and Emiliano is that the lower-left plot shows the portfolio to be equal weighted. We will probably punish that type of pattern and invite you to not rely on maximize_alpha or the optimizer at all. You can see the template algorithm I shared above for how to get to a more sensible weighting scheme. Also, see the tearsheet review of an earlier challenge for more info on what we are looking for: https://www.youtube.com/watch?v=r5FRV5XnY1M

@Thomas. Thanks for the review. I'll remove the optimizer constraints and resubmit.

looks bad over this time span...

Model 2B (alternative ranking).

Model 3A.

Model 3B (alternative ranking).

.

Model 3.1A (updated factors with slightly higher uniqueness score).

Model 4A.

alphaV

Model 5A.

Model 5B (alternative ranking).

Model 4B (alternative ranking).

Model 4.1B

technical

Model 6A

Model 6B (alternative ranking).

Model 7A - Average Daily Turnover: 21.03% (raw, un-smoothed).

Model 7B (alternative ranking) - Average Daily Turnover: 20.85% (raw, un-smoothed).

2005-2019 Model #2

2014-2018 Model #2

Model 1

Model 0.2

Model 7.1A (smoothed). Average Daily Turnover: 10.26%.

Model 7.1B (smoothed). Average Daily Turnover: 10.19%.

Model 8.

Model 9A.

Model 9B (alternative ranking).

Model 10A.

Model 10B (alternative ranking).

Model 11 - Value Factor Composite, each statistically robust (or so I've convinced myself anyway).

Model 12 - Growth Factor Composite, each statistically robust (or so I've convinced myself anyway).

Model 12.1 - Growth Factor Composite (alternative ranking)

Model 2

Strategy #1, looks like the universe is small based on the FS fields used.

Model1:

Model 13A - Three pure Insiders factors (each statistically significant on held out data).

I'm worried that my previous models were possibly overfit, so this and future models have increased focus on robustness, simplicity, and statistical significance.

Model 13B - Same factors; alternative ranking.

Model 14.

Strategy #2

Model 15

Model 3

Model 1

Model: 3A

@Thomas,
Is it ok to use targetweights with some constraints? Graphs looks much nicer.

Model #12

Model #13

Model #14

Model#15

model #16

Model #17

Strategy #3 mixes various sentiments around the market

Model #18

Model #19

Model #20

Model 1

V6 - 2 Factors . Only uses FS Guidance and FS Estimates.

Submission v2

Model 16

My entry 1

Model #21

Model #22

@all: The submissions so far look really great. One pattern however, that seems to run pretty common is one where the exposures are long momentum and short value and volatility. There is nothing wrong with that per se but note that being in a cluster of strategies with very similar exposure patterns will decrease your chances of winning. It also seems that this type of exposure pattern results in lower uniqueness scores.

If you find that pattern in your algorithm, do NOT use the optimizer to get rid of it, but rather realize that your idea might not be as unique as you might have hoped for and see if you can try things more off the beaten path.

Seems like if you have both low turnover and positive returns, that would translate to at least incidental momentum exposure. It's a tautology, no?

Low turnover and low alpha decay will lead to static exposures to factors. If the factor generates a return, it'll also be exposed to returns (momentum). If value generated a negative return over the time period, it will have been short value.

Is there a way around this? Perhaps by shrinking the portfolio to a subset of the universe where the exception to the rule prevails?

S&P is on a bull market. Value is not doing well

Model 1.

Model 2.

Low volatility equity returns

Model 2

Model #3

Model 2

Model 3

Alt

Model4.1

Model #23

Model #4

@Thomas One question and two comments about the tearsheet code.

1. Is the @ operator in compute_uniq()specific to Qunatopian? I have never seen this used in NumPy or Pandas.

2. I don't think you need the line:longs = expos.loc[expos > 0] in get_max_median_position_concentration()and the similar line after it with shorts. I don't see it used anywhere.

3. You can get rid of the annoying red warning when you are fixing the risk loadings bug risk_returns.loc[risk_returns.value.idxmax(), 'value'] = 0 by adding
   factor_returns.is_copy = False prior to the bug-fix line.

Model #5 - Single Value Factor

Model 17A

Model 4

Model 17B

Model 0.3

Model 0.4

Model 1.1

Model 1.2

Model 4 v2 -Alternate weighing

Wow, so many low turnover, high Sharpe ratio, low correlation strategies. It's actually quite discouraging. Unless this is a festival of overfits, congrats everybody for cracking the secrets to the stock market.

Model #24, using the optimizer

Model 1.3

Model 1.4

Single Factor - 1st Submission Model 1

Model A

Submission v2.1

Submission v2.2

Model7

Model8

Model 2 , based on Earnings Expectations

Submission v1.2

Submission v1.1

Modl 5 revised - Nonlinear ML

Model 6 - Nonlinear ML

Long Backtest of Model 6 (For illustration purposes only to show long term consistencies of performance under different market regimes/conditions).
From 01/04/2006 To 01/31/2019

Submission v3

Submission v3.1

Submission v3.2

Submission v4

Submission v4.1

Submission v4.2

Model 18 - Financials specific model.

Model 19

Insider-only

Model #24 One factor

Single Factor 0

Model #25

Model #26

Model #27

Model #28

Model #29

Model #30

Model 12

Model 7 -Nonlinear ML

Model A

Model AA

I have a question regarding the following property:

• The portfolio must hold at least 100 assets in the QTU

My question:
Would it be ok if the portfolio on average holds 140 assets in the QTU and 10 assets outside the QTU?

Model 3

Model 8 - same inputs and ML algo as Model 7 but added another factor derived from a separate ML algo that determines which current state each individual asset is, where state defined as trending, mean reverting or neutral.

Model 8a - same inputs and ML algo as Model 7 but added another factor derived from a separate ML algo that determines which current market regime is at, where regime defined as trending, mean reverting or neutral.

Model 1

Model 2

Strategy 2a

Model 9

Model 10

Alpha Factor #2

another factor

Model 9 - ChaosTheory principles implemented in Nonlinear ML

Alpha #3

Here's the long backtest of Model 9 (01/04/2006 - 02/08/2019). This is an alternative validation process to see if the model in sample performance is consistent on data it has not seen before. It can give some sense of statistical confidence and verifies accuracy of generalization in the face of changing market regimes and conditions.

Model 2.1

Model 2.2

Model 3

Model 4

Model 2.3

Model 2.4

Model 2.5

Model 2.6

Model 2.7

Model 2.8

Model 5

Model 6

Model 7

Model 3.1

Model 3.2

ST BB V

My first submission

Submission v5

Submission v5.1

Submission v5.2

Submission v6

Submission v6.1

Submission v6.2

Submission v7

Submission v7.1

Submission v7.2

Submission v8

Submission v8.1

Submission v8.2

Model15

Model16

Model17

Model14

My first submission

alt 2.0

Model 10

Model 11

Model13

Model18

Model 12

Model19

Model 13

2

3

Submission 4/ 2 factors with data limited to 400-450 Stocks

Model v1, using some insider data.

Model 14

Model A1

Model B1

Model C1

Model D1

Model E1

Model F1

Model G1

sub 5/ Only 1 factor with limited data (not estimates)

Sub 5.1/ the same 1 factor above.

Fund V1

Model #31

Model #32

Model #33

Model #34

Model 16

Model 17

Model 18

Model 19

Model 20

Model #35

Model 4.0

Model 1.1

Model 21

Model 22

Model 23

Model 24

Model #36

Model #37

v1.12

6/ Only 1 factor.

Model 2

Model 1, Version 1

I am not sure what to make of the huge turnover spike around Dec 2015. On the backtest, it is a lot tamer, corresponding to an increase in turnover from 16% to 22%

Model 1

Pension-01-bt7--Insiders+ML
Only up to 02-21-2018 due to Insiders data availability.

v 1.1

Model 2

Submission

Model 1, Version 2

Avg_IRS = 2.41; uniqueness = 97.05; Avg_n_holdings = 279; Avg_turnover = 0.078; v3 Backtest 34

Submission

Strat #8, high turnover

Avg_IRS = 2.48; uniqueness = 97.19; Avg_n_holdings = 402; Avg_turnover = 0.080; v5 Backtest 1

Avg_IRS = 2.51; uniqueness = 96.90; Avg_n_holdings = 616; Avg_turnover = 0.073; v5 Backtest 3

Avg_IRS = 2.53; uniqueness = 97.06; Avg_n_holdings = 737; Avg_turnover = 0.076; v5 Backtest 6

Avg_IRS = 2.54; uniqueness = 97.36; Avg_n_holdings = 725; Avg_turnover = 0.086; v5 Backtest 7

Avg_IRS = 2.55; uniqueness = 97.40; Avg_n_holdings = 765; Avg_turnover = 0.087; v5 Backtest 9

Avg_IRS = 2.57; uniqueness = 97.42; Avg_n_holdings = 795; Avg_turnover = 0.089; v5 Backtest 11

Model 6.1

Model #6

model 1

v 1.13

v 1.14

v 1.15

Model 8

Model 9

Model 10

Model 11

Model 10

Model 3

Model #38

Model #39

Model #40

Full tearsheet

Pension-01-bt27--MultiFactors+ML
Only up to 02-21-2018 due to Insiders data availability.

v1,1

v1.5

v1.3

v1.8

Model 2, Version 1

Submission v9

Submission v10

Model 1 - Values

Model 11 - same inputs as Model 9 but equally weighted factor combination (non ML)

Model #7

Model 20A

Model 21A

Model 4

Model 5

Model 20B (alternative ranking)

Model 21B (alternative ranking)

Model #8

Model 1.0 based on earnings surprise

Model23

Model #9

I_LOVE_CATS NON-LINEAR A1

Model #10

Submission

old strategy

first one

Submission 1 V1

Submission 1 V2

Model 2, Version 2

Model24.1

Model 2

model 1.1

Model 1.2

Submission v11

Submission v12

Model 2 - T Strength

Model A

Model B

model 3

model 4

Submission

Submission v13

Submission v14

Submission v17

Submission v18

Submission

Submission v15

Submission v16

Model 4 V1

Model 19 V1

Model 4 V2

v2 Backtest 10

Model 22

Model C

Model D

Model E

model 5

Model F

Model G

model 6

Model 15 V1

Sub 7

model 7

model 8

model 9

Sub 8

model 10

Model 1

Model 2

Model 3

Model 4

Model 5

Model 6

model 11

RCA Model 1

Model 7

Model 8

Model 9

model 12

Model 10

Sub 9

Model 3.3

Model 3.4

Model 3.5

Sub 10

Model 1.0

Model 1.1

Sub 11

sub 12

Submission

Pension-#2

Sub 13/ 1 lonely factor.

Sub 14

Model 16 V1

Model 15 V1

Model 2

Model 14

2

1.4

Model 4.1

Model 4.2

Model 4.3

model 1.5

Model 14 V2

Model 4.4

Model 4.5

Model 4.6

Model 4.7

Model 4.8

Model 13 V1

Model B

Model C

Model D

Model E

Model F

Model G

Model 16 V2

Model 8 V1

Model 12 V1

Model 11

Model 10

Model 8

Model 7

Model 5

Model 3 V2

ES bt2

EU3 bt2

CA probability weighted

Old factor without constraints

EG probability weighted

EW

model 3

Submission 2

Submission 1 (edited)

Model 6

Model 11

Model 1

Model 2

Model 3

Model 4

Sub15/ 1 factor.

model 4

Sub 16

Model 4.9

Model 5.0

Submission I

Sub 17

Sub 18

Sub 18.1

Sub 18.2

Thanks everyone for submitting, the challenge is now officially closed!

Model 3: Tech IND cross. Sorry I was late.

Model 4 - TECH IND all.

Hi everyone!

There is an update for this challenge posted here. Join our LIVE winner announcement and tearsheet review webinar here on Monday, April 6 at 2 PM EST.

Congrats once again to our winners!

• Kyle M
  
• Vedran Rusman
  
• Oleg
  
• Oleg
  
• Vedran Rusman
  
• Joakim Arvidsson (Cream Mongoose)
  
• Indigo Monkey
  
• Joakim Arvidsson (Cream Mongoose)
  
• Daniele Carabini
  
• Joakim Arvidsson (Cream Mongoose)
  
• Bjarke Riis Larsen
  
• Viridian Hawk
  
• Vladimir
  
• Leo M
  
• Neeraj Salodia
  

Your prizes are being processed. You should receive an email from our payment service soon.