@dark-pidgeot said in Q20 contest results:

@support said in Q20 contest results:

That is also a good observation. In general, we would like a long-only system with low turnover to beat the Nasdaq index for example. But still, we see a lot of problems with hand-picking stocks and survivorship-bias affected systems.

if I understand correctly

if I initially select 10 stocks, and even if I have a good sharpe ratio, my strategy can make long and short positions. I could be disqualified

That's my understanding too. If you initially select 10 stocks, that already implies survivorship-bias, and is in fact "forward looking" because in for instance year 2010 you could not foresee which 10 stocks would still exist in the index in 2024.

@support Thank you for clarification!

@vyacheslav_b

This was the last page. If I set it to show 1000 rows per page I still only see results from 2023-05-15.

Did I miss something?

Hi,

I was playing around with strategies and found that the evaluation only starts from one year back? I remember previously it started from 2016-01-01. This kinda breaks my submission.

I'd love to know if this is intentional. Thanks

@vyacheslav_b Thank you! I'll have a look

Hi, I'd like to access to access both market and index data in strategy(). Would the following code work? It worked locally but when submitted the "checking" progress does not move at all.

# load data
data = qndata.stocks.load_ndx_data(min_date="2005-01-01")

def strategy(data):
    # get index data
    index_name = 'NDX'
    index_data = qndata.index.load_data(assets=[index_name], min_date='2005-01-01', forward_order=True)
    index_data = xr.align(index_data.sel(asset=index_name), data.isel(field=0), join='right')[0]
    ndx_values = index_data.values

I also tried a load_data function where I load both datasets and use load_data=load_data in qnbt.backtest() , but it quickly leads to a memory leak.

@support

Thank you, I noticed some changes on the rules so I have to reimplement the strategy anyway.

Hi I have a strategy in q20 but I want to reuse it in q21.

I reuploaded it as # 16543405 but it says

"The strategy has a high correlation with standard templates or your strategy from the past competition:
#15136749 money-launderer-q20-test01-b - 1.00 (IS null)"

This is the one I want to reuse. Would it be possible?

Best regards

@vyacheslav_b

Thanks for the reply. I know "market cap = #shares * price". The real issue is that #shares is not available for some stocks (e.g. META) so I have to rely on other data sources like yahoo finance.

@support

Do you have an update? Thanks!

@support I have 3 more questions regarding adjusted close.

(1) I have not checked the source code but I assume your backtest / online evaluation is based on split- and divident-adjusted close prices?

(2) In qnt.qndata.stocks only close is provided. I checked it is split-adjusted properly but not dividend-adjusted (see attached screenshot). Should we retroactively compute the adjusted close based on divs? It would be nice to have it in the data so we don't have to compute it every time.

(3) The split_cumprod column is always 1. What does it mean? It's probably not important.

@support Thank you, looking forward to your update.