Research Hypotheses

In this chapter, Section One will develop the hypotheses for this study, which are based on two economic models. Section Two will present the data selection process.

Section Three will discuss the research methodology and design, and Section Four will examine the empirical models and variables.

3.1 Research Hypotheses

3.1.1 The Production Smoothing Model

The production smoothing model is one of the most widely studied models of inventory in economic literature (Blinder 1986). A necessary motive for a company to smooth production is that demand varies through time. If there is a random element to demand, a company may decide to smooth production and treat inventories as a buffer stock. Therefore, a firm is said to smooth production if the variance of production is less than the variance of sales.

The information structure of the production smoothing model presumes that both cost shock and demand shock would affect production decisions. According to Guido Lorenzoni (2006), demand shock is a sudden event that causes a shift in consumer expectations, which increases or decreases demand for goods or services temporarily, while cost shock is an event that causes a sudden increase of decrease of production costs. The production smoothing model assumes that managers can observe cost shock and part of demand shock before choosing its level of production, price, and expected sales. After these decisions are made, the rest of the demand shock is observed and actual sales are determined. The inventory levels for next period then follow and modify the prior production decision.

Consequently, we can see that when the production is smoothed, the resulting

‧

inventory levels represent management’s expectations about future demand and cost structures, which may also include management’s private information. As a result, inventory levels can be positive leading indicators of future sales when interpreting financial statements. In addition, unless competitive forces totally eliminate any impact of sales changes upon earnings, inventory levels should also be positive leading indicators of future earnings.

Under LIFO, the changes in inventory mostly represent the changes in inventory volume, while under IFRS, the changes in inventory represent the changes in both inventory volumes and current costs. It is because under LIFO, the items remaining in inventory are recognized as if they were the oldest, so the inventory costs remain the same throughout the year. Thus, any change in inventory levels reflects the inventory volume change. Under IFRS, because the items in inventory are measured by

inventory’s current cost, the changes in inventory levels may result from the changes in costs or volume.

This study further assumes that when inventory volume is the only factor that affects inventory levels, inventory levels will be stronger indicators of future sales and earnings. Therefore, this study assumes that inventory levels reported under LIFO method should be stronger positive indicators of future sales and earnings than inventory levels reported under IFRS method.

Hypothesis 1:

Under LIFO method, inventory levels are stronger positive indicators of future sales than under IFRS method.

Under LIFO method, inventory levels are stronger positive indicators of future earnings than under IFRS method.

‧

3.1.2 The Stockout Model

The stockout model is one of the inventory models that are more consistent with existing data (e.g., Kahn [1987]). In the stockout model, if actual sales are less than the available stock, the company may carry the remainder into the next period as inventory. If, on the other hand, actual sales are more than the available stock and the company ―stocks out,‖ it generates losses, and if a buyer is willing to let the company sell the product in next period at this period’s price, the company will occur a backlog in next period. As a result, when making production decision, a company must weigh against the possibility of stockout and the possibility of holding excessive inventory.

According to Kahn, under a stockout situation, a company’s sales consist of backlogged sales from previous periods and current demand from this period, so current demand is only partially reflected in current sales; the remainder of current demand is reflected in the frequency of stockouts. A low inventory level indicates a potentially high frequency of stockouts, which further indicates higher level of

demand and sales. On the other hand, a high inventory level indicated a lower level of sales. Consequently, inventory levels are inversely related to future sales. In addition, inventory levels are also leading negative indicators of future earnings, because the lower sales may lead to lower margins, and higher inventory levels lead to higher inventory holding costs.

The stockout model can rationalize the violations of the production smoothing model because it suggests that production can be more variable than sales. Two situations may lead to production counter-smoothing. First, because backlogs may shift sales away from large unexpected demand, while production still responds to previous period’s excess demand, the variance of production is larger than the variance of sales. Second, when demand shock occurs, it changes the ending

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

17

inventory and the expectations about future demand, which increases or decrease optimal production, so the variance of production is larger than the variance of sales.

Under LIFO, the changes in inventory represent the changes in inventory volume, while under IFRS, the changes in inventory represent the changes in both inventory volumes and current costs. This study further assumes that when inventory volume is the only factor that affects inventory levels, inventory levels will be stronger

indicators of future sales and earnings. Therefore, this study assumes that inventory levels reported under LIFO method should be stronger positive indicators of future sales and earnings than inventory levels reported under IFRS method.

Hypothesis 2

Under LIFO method, inventory levels are stronger negative indicators of future sales than under IFRS method.

Under LIFO method, inventory levels are stronger negative indicators of future earnings than under IFRS method.

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

18