The next method is sometimes called the naive method, but it's not a bad method once you think about it. Again, we have this notation y^T+h|T plus h. So h is the number of periods ahead you're looking. If you want to simplify, as you think about it, just set h equal to 1. So it's the next period, and y hat means what's my forecast for T plus 1? What's my forecast for tomorrow given I have today's value. I only have today's value In this case, my best guess for tomorrow's value is today's value. That's not a bad method if you think about it. If you think about stack prices, you can think about some corporate stock has a price of a $100 per share, and it's not fluctuating randomly from day to day. It's going from 100 to 97 to 42 to one million and seventeen, down to three, it's not some random draw from a normal distribution. It is dependent on the previous phase. So if you have no other information, you know nothing else about the stock, and someone were to ask you, what is the best guess of the price of a stock X? If you know today's price is 100, your best guess would be tomorrow. It's going to be 100. Common problems might say it's a 100 give or take. But your best guess is essentially 100, and that's what this naive method represents. So your best guess for tomorrow, your forecast for tomorrow t plus 1 is equal to the value of today. So let's look at r. So in r, the command is naive. I'll be still using the same beer dataset. Let's run this little piece of code here, and you can see my forecast is a 153, which is equal to, let's look at the beer dataset again. The last value here August '95. So my best guess for September of '95 is the value of 153, which is what we know was produced in August of 95. Let me clear the screen. We just saw naive beer, one which is one period ahead. Let's look five periods ahead or up to five periods ahead. There you have the table. Again, the forecast for each of these time periods is the last known value, which is 153. This command also gives you a confidence interval of 80 and 95 percent confidence intervals. Another way to talk about a naive forecasts, it's also equivalent to a random walk. So here's the random walk come in. Rwf 5, there it is. It has the same data. A random walk basically means I start at some point, and I have some arbitrary movement plus or minus in some direction, and I get to the next point, and then I have another arbitrary movement plus or minus and I get to the next point, and over time, you will bounce around. We'll talk a little more about that later on when we talk about the ARIMA models.