Welcome to this discussion thread. Feel free to ask questions or discuss things related to this section.
@filip, is the “time window” for a simple moving average (SMA) always N days to the past?
Does it make sense to have a window other than 20 days?
Does it make sense for some past days to have a window of [t - N/2 … t + N/2] considering some “future days”. Of course, it is not applicable for “today” when your trades are supposed to happen.
I just wonder “why SMA is calculated in a way you’ve explained it”. Maybe it would be covered in upcoming videos.
@filip , the “support” video link takes here which might be an issue.
My “support” related question
Does this metric make sense on short-term trends?
You’ve showcased it on the long-term (zoomed out) graph and have not mentioned “support” meaning for short-term.
I guess, that’s because it makes no sense. Just wanted to double check and clarify. So I’ve asked this question.
Yes it does apply in short terms trends. But generally not as strongly. Supports are usually stronger in long term timeframes. Not as reliable for short term trends.
I don’t really understand the question or your suggestion. Moving average is calculated N days backwards. But it’s done for every day in the set of days in the chart. So it forms a moving line. Maybe you can clarify what you mean.
Maybe you can clarify what you mean.
I was asking whether it makes sense calculating moving average by
summing N/2 days from the "past" and N/2 days of the "future" for each data point.
When I was studying statistics in the university we did such calculations for random “time series”. Unfortunately I don’t remember the context of that data, though. I guess, we got that from some “probablistic” random number generators we had implemented in the same course before.
Just wondering if it makes sense to calculate moving average in this way for financial timeline data.
Not that I know of. The problem is that you can’t look into the future. So that would only be a historical indicator. Because if I use your formula, I can’t get the moving average for today since I don’t have the data for N/2 days ahead.
Greetings Filip, I have heard something about a 20 day moving average line crossing a 200 day moving average line and vise versa. Is this a change in trend? What does it mean? Thank you for the excellent information so far.
Yes, usually it’s MA50 crossing MA200 that’s called the Golden Cross or the Death Cross. Depending on where it’s crossed from. It generally indicates that the longer term is starting to switch. You can read more about it here: https://www.investopedia.com/terms/d/deathcross.asp
This is an excellent article that explains it well. It also explains that when one sees something like this that one should wait several weeks for confirmation. Thank you for the reference.
Yes that’s absolutely correct. Like everything else, these crossings can give false signals as well.
does anyone know which software Filip is using during the videos? Because it would be very useful to try it as the same time and play around with the different tools.
I found this one but it’s not the same.
I use tradingview.com
Thanks for the answer
Thank you for the great content.
Your example shows the MA20, what is the best MA or EMA period to look at when trying to figure out the switch between the bull run trend and the beginning of the bear market?
The 2017 crash after the peak occurred in just a few days, let’s say a week.
Does it means we should mainly look at let’s say the MA/EMA,7 to anticipate such a market switch on a daily scale?
There is no easy way of determining which MA/EMA is “the best”. As I think I said in the video, the longer MA’s are obviously more correct in that they have less false signals, but they are very delayed indicators. Shorter MA’s are more prone to false signals but will switch faster.
It all depends on how you are going to implement and use it. You need to backtest those ideas and see what works for different instruments and markets.
Thank you for the clarification.
In general, how many consequetive higher highs or lower lows you need to confirm a trend on a given time frame?
Unfortunately there is no easy answer. The best I can tell you is that it’s all relative. A trend that has lasted for 10 years is a very strong trend, and the consequences of breaking that trend will be bigger.
A trend that has made 1 or 2 higher highs/lows could be the start of a trend but you can never get a “confirmation” since no one know if the trend will break in the next hour. So you need to treat the trends according to how strong they are, relatively speaking. So if a trend has made 20 higher highs on the daily timeframe, you should pay more attention to that trend and its trendlines compared to a trend that just made 5 higher highs. You get my point?
Yes, makes sense now - to see the trend in the larger context rather than on an isolated time frame. Also, I was palying around with the indicators and oscillators and it seems like a good idea to test the trend’s strogness using MA, for instance?