# Safety Stock Explained

Safety stock is an important and interesting concept in inventory management. However, not every procurement professionals can explain this concept clearly. Some mistake safety stock level with reorder level. Even CIPS Study guide L4M7 does not provide any formula to calculate or at least estimate the quantity of safety stock. Fortunately, the resources on Internet are plenty and free. We have done some research and in this post, we will explain the use of safety stock, formulas to calculate the safety stock level and the origins of such formulas.

### The use of safety stock

Safety stock is also known as buffer stock. As this name suggests, this type of stock provides some kind of ‘buffer’, which means safety stock will help the business to reduce the shocks induced by volatile demand or disruption on the supply chain. In other words, safety stock will reduces the probability of stockouts. The following graph would explain the reason why an organisation should have safety stock:*Figure 1: Safety stock graph*

As it is only the buffer against uncertainty, safety stock level should be equal to the deviation of demand or replenishment time. For example, imagine that you run an apparel shop which normally sells 100 shirts per week. This is only the average number. In high demand seasons, you may sell up 220 shirts per week, while in low demand season, you may try your best but the sale is only 40 shirts per week. The differences may be great and you need a safety stock to guard against such volatility. But how to calculate it?

### Formulas of safety stock

There are many articles on the formulas of safety stock as well as their advantages and disadvantages. In this section, I only list the formulas.

#### Method 1: Old-fashioned way

*Figure 2: Old-fashioned Safety Stock formula (Source: https://abcsupplychain.com/)*

Method 2: Max – Average

**(maximum sale x maximum lead time) – (average sale x average lead time)**

#### Before approaching to other 4 methods using Normal Distribution, I would like to remind you about service level. Service level is the measure of how many times your business can provide the goods to customers without delays. Normally the service level is expressed in percentage value (i.e. 80% means that 4 out of 5 times you deliver the goods to your customers without delay). Reaching 100% service level is impossible, so people often set out the service level target and use Z-score as service level coefficient. In Microsoft Excel, you can find Z-score by the function NORM.S.INV.

Method 3: Normal Distribution with uncertainty about the demand

**Safety Stock = Ζ × √ ^{PC}⁄_{T} × σ_{D}**

Where: Z = Z-score

PC = performance cycle, another term for total lead time

T = time increment used for calculating standard deviation of demand

σ_{D}** _{ }**= standard deviation of demand.

#### Method 4: Normal distribution with uncertainty on the lead time

**Safety stock = Z × σLT × Davg**

Where: Z = Z-score

σLT = standard deviation of lead time

Davg = average demand.

#### Method 5: Normal distribution with uncertainty on-demand and independent lead time

**Safety stock=Z×√((PC/T × σD^2 )+(σLT × Davg )^2 )**

Where: Z = Z-score

PC = performance cycle, another term for total lead time

T = time increment used for calculating standard deviation of demand

σD** **= standard deviation of demand.

σLT = standard deviation of lead time

Davg = average demand.

Method 6: Normal distribution with uncertainty on demand and dependent lead time

**Safety Stock = (Z × σLT × Davg) + (Ζ × √ ^{PC}⁄_{T} × σD)**

Where: Z = Z-score

PC = performance cycle, another term for total lead time

T = time increment used for calculating standard deviation of demand

σD** **= standard deviation of demand.

σLT = standard deviation of lead time

Davg = average demand.

### New (and perhaps more accurate) Techniques to Calculate the Safety Stock Level

In 2017, three researchers from Netherlands and the UK developed a new technique to calculate the safety stock level. According to their research, traditional approaches can lead to safety stocks that are up to 30 percent too low and service levels that are up to 10 percent below the target. You may get access their research paper here.