RDP 8904: Changes in the Behaviour of Banks and their Implications for Financial Aggregates 2. Removal of Interest Rate Controls and the Shift to Liability Management

(a) Interest rate controls and asset management

Up to the early 1970s, there were controls on the interest rates banks could pay on deposits, on the interest rates they could charge on loans, and on the maturity of deposits they could raise. (See the Appendix for a description of controls in force at that time.) In these circumstances, banks behaved largely as asset managers. They accepted passively whatever deposits came their way, since controls over interest rates limited their scope to go out into the market and compete for deposits. They then decided how to allocate these funds among various assets.

On the asset side of their balance sheets, banks adjusted to short run changes in inflows of deposits by movements in their holdings of liquid assets and Commonwealth Government securities (LGS).[1] During times of large inflows of deposits, LGS assets were run up and, when deposit inflows slowed or outflows occurred, LGS assets were run down. On average, banks held a large buffer of LGS in excess of minimum requirements. [2]

Fluctuations in holdings of LGS helped to shield advances from variations in deposit flows. However, in the event of a tightening in financial conditions, banks eventually were forced to adjust their advances, as they could not competitively bid for deposits to replenish their liquid assets. Credit rationing on a non-price basis was an important part of the adjustment process.

In these circumstances, a tightening of policy tended to have a substantial and relatively fast effect on the growth of bank deposits. There were two main channels through which policy could be tightened. One was the LGS/SRD mechanism. By raising the SRD ratio, banks could be forced to cut back on holdings of other assets. In the short run, this usually involved relinquishing LGS but if this brought them close to the minimum LGS ratio, they would be forced to restrain lending. This in turn restrained deposits. Alternatively, policy could be tightened by more active sales of CGS (including issues of Australian Savings Bonds). This forced up interest rates, not only on CGS but also more generally. To the extent that sales were to non-banks, it also directly attracted deposits out of the banking system as non-banks paid for their securities. Banks were restricted in their ability to compete for deposits because of controls over the interest rates they could pay.

Graph 1 illustrates the above adjustment process by reference to the late 1960s and first half of the 1970s, a period during which banks behaved overwhelmingly as asset managers. [3] While interest rates on CDs were freed in 1973, banks did not move immediately to make full use of this freedom; it took some time to adapt behaviour, and the CD market initially was not very big. It was not until the second half of the 1970s that clearer signs of liability management began to emerge.


The top panel of the graph shows two interest rates: the 90-day bank bill yield (as an indicator of market rates), and the interest rate on trading bank fixed deposits. The graph covers two periods of policy tightening, 1970 and 1974. In the first period, the bank bill rate rose by about 4 percentage points, but the interest rate on deposits rose only a little, as it was constrained by a ceiling. The gap between market rates and bank deposit rates closed in 1971 as market rates fell. The experience was repeated in 1974: the bill rate rose sharply, but the rate on fixed deposits was again limited by the ceiling.

The second panel shows the growth of trading bank deposits and advances. In each period of tightening, the growth of bank deposits slowed noticeably. The response was also fairly quick. Growth in deposits turned down before interest rates reached their peak. The slowing in advances took longer to emerge; turning points for advances were at least one quarter behind those for deposits.

The bottom panel illustrates trading banks' holdings of liquid assets in excess of minimum requirements. Comparing this with the middle panel, it can be seen that banks met slow-downs in deposits initially by running down their excess holdings of LGS, which had been built up in periods of sustained deposit growth. Banks' excess holdings of LGS assets fell to only about 2 per cent of deposits in both 1970 and 1974, compared with average holdings during this period of about 7 per cent of deposits.

(b) The move to liability management

Two major steps in the removal of interest rate controls on the banking sector were taken in September 1973, when interest rates payable on certificates of deposit were freed, and in December 1980 when virtually all other controls on bank deposit interest rates were abolished. Restrictions on the minimum maturity of CDs and fixed deposits remained until August 1984.

These changes mainly benefited trading banks, allowing them to move from passive acceptance of deposits to active management of deposits. While interest rate controls on savings bank deposits were also removed, savings banks continued to be constrained in the interest rates they could pay on deposits by the continuation of a ceiling on interest rates they could charge for their housing loans. (This issue is discussed in Section 2(c) below.) Following the removal of controls, trading banks could set competitive rates on CDs and fixed deposits, enabling them to tailor the inflow of deposits to match the demand for loans. Fixed deposits and CDs began to account for an increasing share of deposits.

This is illustrated in Graph 2. The proportion of total trading bank deposits accounted for by fixed deposits and CDs rose from a little over 40 per cent in the late 1960s to around 70 per cent in the mid 1980s. The major increases took place soon after 1973, when interest rates on CDs were freed, and after 1980 when controls on fixed deposits were removed.[4]

(per cent of total)

With greater freedom to set interest rates on deposits, banks moved towards liability management. This is illustrated in Graph 3 which shows the same series as Graph 1 over the period from 1975 to 1988. The dominance of liability management is most pronounced in the 1980s. The effects of deregulation of bank interest rates is clear from the top panel of the graph. After 1980, rates on fixed deposits moved much more closely in line with bank bill rates. The yield on CDs was virtually identical to that of bank bills. There were no cases after 1980 where bank deposit interest rates were left well behind by movements in other short-term interest rates.


From the second panel, it can be seen that from the late 1970s, the growth rates of bank deposits and advances moved together much more closely than earlier.[5] With controls over interest rates removed, trading banks could manage deposit flows by varying interest rates on CDs and fixed deposits, so as to keep deposits in line with the demand for advances.

The responsiveness of trading bank deposits to a tightening in financial conditions changed after the first half of the 1970s. Whereas in the first half of the 1970s they responded quickly (and well before advances) to a change in interest rates, this was no longer the case in later periods. The policy tightening over 1981/82 is a good example; despite the sharp rise in interest rates, growth of deposits did not slow until lending slowed.

The third panel shows that banks' excess holdings of LGS have been much lower, and much more stable, in the 1980s. With the ability to attract deposits as required, it was no longer necessary, and certainly not profitable, to maintain excess liquid assets as a buffer against a tightening of financial conditions. By the early 1980s, excess holdings were averaging less than 2 per cent of deposits, compared with an average of about 7 per cent in the first half of the 1970s. In the two major tightenings of financial conditions in the first half of the 1980s there was little, if any, change in the excess LGS ratio, whereas the tightenings in the first half of the 1970s produced major falls.

(c) Savings banks – continued asset management

The experience of savings banks is in sharp contrast to that of trading banks described above.

The structure of savings bank balance sheets remained constrained by regulations for much longer than that of trading banks. Savings banks did not have the outlet provided to trading banks by the certificate of deposit.

Prior to August 1982, savings banks were prohibited from accepting deposits from trading and profit-making bodies and prior to August 1984 could only offer small fixed deposits. These restrictions effectively barred access to wholesale deposits.

Even more importantly, savings banks were restricted in setting deposit rates by the continued regulation of the housing loan rate. The mortgage rate (for loans of less than $100,000 to owner-occupiers) was subject to a ceiling until April 1986, and loans outstanding approved before that date are still subject to that ceiling.

These restrictions meant that savings banks often did not have the freedom to set competitive deposit rates or otherwise significantly influence the inflow of their deposits until comparatively recently. Without this ability, savings bank deposits remained very sensitive to changes in market interest rates. When rates rose, flows of funds into savings banks quickly dried up. By the same token, when market interest rates were falling, savings banks were generally slow in reducing their deposit interest rates, preferring instead to take in deposits to rebuild liquidity.

This is illustrated in Graph 4, which shows the growth of trading bank and savings bank deposits and the 90-day bank bill rate.


In the first half of the period shown, the experiences of savings banks and trading banks were similar – deposit growth falling when market interest rates rose, and picking up when market rates fell. However, whereas for trading banks this relationship broke down in the 1980s, it persisted for savings banks, with the result that by the early 1980s growth in trading bank and savings bank deposits were following opposite patterns.

For example, as financial conditions tightened in 1984 and 1985, growth of savings bank deposits fell sharply, with funds attracted away by the more aggressive behaviour of trading banks. In the periods of weakening economic activity and falling interest rates in 1983 and 1986, trading bank deposits slowed in line with the growth of their advances, but growth in savings bank deposits picked up.

(d) Implications for monetary aggregates

The implications for M3 of these changes in behaviour of trading and savings banks can be seen in Graph 5. It plots annual growth rates of trading bank deposits, savings bank deposits, and M3.

(Growth in 4 quarters ended)

In the early 1970s, all three lines followed roughly the same path. If financial conditions were easy (e.g. as in 1972/73) both savings banks and trading banks experienced large inflows of deposits and growth of M3 increased. When conditions tightened, as they did in 1974 for example, both groups of banks experienced slower growth of deposits and growth of M3 fell.

In this early period, a tightening in monetary policy tended to have a fairly quick and predictable effect on M3. Growth started to fall in the same quarter that interest rates rose. The response of M3 preceded that of lending, with banks initially adjusting by running down their holdings of LGS, and later adjusting loans and advances. This behaviour is consistent with the finding by Bullock, Morris and Stevens (1988) that M3 was a leading indicator of economic activity in this period, while advances were co-incident.

By the early 1980s, the responsiveness of M3 to changes in financial conditions was very much reduced. With trading banks having increased ability to compete for deposits, they were able to maintain inflows until the demand for credit fell. M3 therefore responded to changes in financial conditions only when, and to the extent that, the demand for bank credit did. Its response was therefore both slower and smaller than was the case in the early 1970s, when changes in M3 were related not only to changes in credit but also to changes in holdings of LGS assets.

The responsiveness of M3 was further reduced by the fact that savings bank deposits followed a pattern opposite to that of trading bank deposits. As noted above, this was because savings banks were still constrained by controls on interest rates and were still operating as asset managers. As demand for advances increased, trading banks bid more aggressively for deposits to fund these advances. This did not fully reflect in M3 as some of the deposits were attracted out of savings banks, who were forced to run down their liquidity. Conversely, as demand for advances fell, trading banks slowed their bidding for deposits. Savings banks, however, were willing to accept large inflows to rebuild their liquidity. Again, therefore, the effects on M3 of a slowing in demand for advances was muted.

The experience in 1982/83 provides a good illustration. As noted in Bullock, Morris and Stevens, M3 growth did not slow much during the recession of 1982/83; the annual growth rate was around 12 per cent in 1981 when economic activity peaked, and despite a sharp weakening in output over the following year or two (the most pronounced in the post-war period), growth of M3 slowed only marginally to reach a low of about 10.5 per cent in 1982/83. Throughout this period, growth of M3 generally remained above the targets in place at that time.

The patterns of trading bank and savings bank deposits illustrate why this was so. In 1981, when economic activity was strong, growth in trading bank deposits had been running at an annual rate of about 14 per cent, broadly in line with the growth rate of advances. Then, between mid 1982 and 1983, growth in trading bank advances and deposits slowed noticeably as economic activity weakened.

Savings bank deposits, however, followed a very different pattern. Growth had been quite low in 1981 – around 8 per cent, roughly half the rate of increase in trading bank deposits. Then, during late 1982 and 1983, as the economy slowed and market interest rates fell, growth in savings bank deposits picked up sharply. The annual growth rate reached nearly 25 per cent in late 1983, at a time when growth in trading bank deposits had fallen to 5 per cent.

During the next economic cycle – the strengthening in the economy over 1984 and the weakening over 1986 – trading and savings bank deposits continued to follow opposite patterns, with M3 following a middle course, largely unresponsive to economic conditions.[6]

The interest rate ceiling on savings banks' new housing loans was removed in 1986, and this appears to be starting to have some effect on savings bank behaviour. The end of the distinction between trading and of savings banks could also be expected to change the behaviour of M3.

Aggregates narrower than M3, such as M1, were largely unaffected by this change in trading bank behaviour from asset management to liability management. Because current deposits, which form the bulk of M1, remained largely non-interest bearing until fairly recently, M1 continued to be responsive to changes in interest rates. Bullock, Morris and Stevens noted that M1 had a stable relationship with interest rates up until recently. It has increased faster in recent years than might have been expected on the basis of past relationships. This has coincided with an increase in the proportion of current deposits bearing interest.

This can be seen in Graph 6. The proportion of current deposits bearing interest has risen from a little over 10 per cent in the early 1980s (a figure that had not changed much over the previous 15 years) to 30 per cent in 1987. The trigger for this rise was the removal in August 1984 of regulations restricting the payment of interest on current accounts. Later, an additional factor was the general fall in the level of interest rates over 1987, which tended to encourage the holding of current deposits.

(per cent of total trading bank current deposits)


Of course, major changes in holdings of LGS assets usually meant changes in holdings of Commonwealth Government securities (CGS). LGS assets which earnt no interest, such as notes and coin and exchange settlement funds, were always kept to a minimum. [1]

The minimum requirement was set by the “LGS convention”, under which banks agreed to keep not less than a certain proportion of their depositors' funds in the form of LGS assets. [2]

All figures shown in the graphs are quarterly averages. Unless otherwise specified, throughout this paper figures for banks have been adjusted for the establishment of new banks, which may have caused the switching of assets and liabilities from non-bank financial intermediaries (NBFIs) to banks. [3]

The increasing importance of these deposits has been partly reversed in recent years when current deposits of banks have risen sharply (see Section 2(d)). [4]

Over 1986 and 1987, a large gap emerged between the growth of bank deposits and that of bank advances. Rather than being a return to the behaviour of the early 1970s, this was a further refinement of liability management, involving the use of non-deposit liabilities. This is discussed in more detail in Section 4 below. [5]

A mild cycle is apparent in M3 over 1985 and 1986 but, as shown in Section 3 below, part of this would have been due to the increased competitiveness of banks vis-a-vis non-bank intermediaries following effects of further deregulation. [6]