The leverage in this algo starts out over 800 and then remains over 20 for the better part of the backtest. Not even Warren Buffet could get a broker to extend them that type of margin. Retail investors are generally limited to a leverage of 2x.

The algo doesn't buy and hold an equally weighted portfolio as intended. The weight shouldn't be specified as 50 but rather something like this

    # buy them all on the first day, then stop  
    # equally weight each stock  
    weight = 1 / len(context.stocks)

    if not context.bought:  
        for stock in context.stocks:  
            order_target_percent(stock, weight)

    context.bought = True  

Attached is an algo which is more representative of a 'buy and hold' strategy on some hand picked stocks (once we knew they were going to be great) .

Uhhh your algorithm gets a net return of 27%, mine has over 600000%, I didn't have to think I'd have to explain the difference to someone with an interest in quantitative trading...

@Thomas,
The Quantopian software is a simulator, which if fed unrealistic parameters will produce unrealistic results that won't verify with real life.

Even if fed "realistic" parameters, it's(the simulator's) out-of-sample(forward-in-time) performance might not match what happens in real-time live-trading(real-life).

In the case of your original algorithm, you'd have to put in your $1k, and get someone to loan you $80k(if as Dan says, it starts out with leverage of 80x) to start off your live trading. If you lose enough so that your large-leverage loan is called...say at -$40k..., you'd face the leveraged version of "gambler's ruin", and owe someone $40k...perhaps in the first day of live trading! Unfortunately, in real life, one of the unintended consequences of risk can be:

https://www.forbes.com/sites/sergeiklebnikov/2020/06/17/20-year-old-robinhood-customer-dies-by-suicide-after-seeing-a-730000-negative-balance/#4e7769111638

On the other hand, the "order_target_percentage()" command and documentation is always tricky, and I've been fooled by it more than once, even when trying to use it in live-trading. I just don't use it anymore except when doing paper-trading research.

alan