## Stop-Losses and Take-Profits

Having Stop-Loss Orders (S/L) and Take-Profit Orders (T/P) allows you to compute a Risk/Reward Ratio for you trades. If you’re trading at the weekly level, these types of orders might let you ride upward trends more reliably than if you did it intraday. Volatile assets might hit a T/P target way earlier than you’d hope, which might be a good thing if you want to dip in and out. If you have a higher threshold that you don’t mind waiting for, then a weekly trading horizon might be enough.

A fixed Stop-Loss is a good idea unless you’re shorting. We know that the higher the loss, the harder it is to recover from it. The Take-Profit could be set by using an asset’s return standard deviation, or the median positive return over a period. Both of these statistics might give you a conservative expectation of where to get out of a trade and don’t require you to wildly guess on where the price action might go.

library(dplyr)

values <- c(344, 345, 321, 367, 456, 312, 250, 323, 410, 405, 398)
data <- data.frame(values)
data <- data %>%
mutate(pct_change = (values / lag(values) - 1))

plot(data$values, main = "Price Action", xlab = "week number", ylab = "price") standard_deviation <- sd(data$pct_change, na.rm = T)
positive_median_change <- median(subset(data, pct_change > 0)$pct_change, na.rm = T) standard_deviation ## [1] 0.2028522 positive_median_change ## [1] 0.2425068 In this hypothetical case, the standard deviation and median positive returns are reasonably close. The time window of your observations is essential. A larger window allows for more data and more reliable statistics. In this scenario, you could choose to put your take profit at $$\approx20\%$$ or $$\approx24\%$$ of the latest price. tail(values, 1) + (tail(values, 1) * standard_deviation) ## [1] 478.7352 tail(values, 1) + (tail(values, 1) * positive_median_change) ## [1] 494.5177 It might look like a big jump, but if you inspect the returns for the period, you can see that the price reached a positive return exceeding $$20\%$$ three times over the eleven weeks in the dataset. That’s $$\approx27\%$$ of the time. data$pct_change
##  [1]           NA  0.002906977 -0.069565217  0.143302181  0.242506812
##  [6] -0.315789474 -0.198717949  0.292000000  0.269349845 -0.012195122
## [11] -0.017283951

## Long-term investing

For long-term strategies (buy-and-hold investing), you could look at large-cap companies. The price action of these companies is relatively stable and “boring.” However, good large-cap corporations increase in value, slowly, but they do. Many of them also tend to have reliable dividends. You can adapt the strategy of buying many shares and rely on:

1. their value slowly increasing over the months (or years).

2. the dividend income to offset some of the losses you incur when trading more aggressively.